13jay Chotaliya

Vidyabharati International Interdisciplinary Research Journal ISSN: 2319-4979
ASM: ADAPTIVE SEQUENTIAL MODELLING BASED RECOMMENDATION SYSTEM
S. Singh1* and K. Srivastava2

1
D.J .Sanghvi College of Engineering, Mumbai
2
Department of Computer Engineering, D. J .Sanghvi College of Engineering, Mumbai
_______________________________________________________________________________________
ABSTRACT
In the last few decades, there is a huge increase in the amount of information available online. This causes information
overload problems making it complex for the users to make decisions. The recommender system plays a vital role in
every individual's life and helps them in taking decisions without exploring physically. Broadly there are two types of
recommender systems: Content-based and Collaborative filtering. Content-based approaches depend on the user’s
previous history. Collaborative filtering finds various hidden patterns by finding a similar group of users. Traditional
recommender system algorithms face challenges to provide an accurate result when the user’s interest changes over
time. In this research work, a deep learning-based recurrent neural network-based model (RNN) is presented which
captures changing preferences of users on the movie Len’s dataset. The comparison of traditional and proposed
methods based on Root Mean Square Error (RMSE) and Mean absolute error (MAE) is discussed.
________________________________________________________________________________
Index Terms: Recommender system, root mean square error, matrix factorization, k nearest neighbor,
single value decomposition, Deep learning, recurrent neural network (RNN).
Introduction recommendations. These filtering methods

takes two types of feedback. Explicit feedback
Recommender systems have changed the way
which is collected from the users as their
people interact with many services. User’s
previous choices. Implicit feedbacks are
feedbacks and evolving technologies are used
collected by browsing history.
to provide personalized suggestions to the
Even though traditional recommender systems
users for better experience. It considers
are used most widely, they do not perform well
independent flow of information provided by
when user’s interest get changed. For example,
user by considering the user’s feedback.
a user watches some action movies before and
Recommender system are used to provide
then he watched some horror Movies.
recommendations based on user’s choice for a
Traditional recommender system will provide
particular product which user might like or
new movie recommendations based on action
dislike. In this era of internet recommender
movies only, since user’s preference has
system plays a great role in influencing
changed afterwards. In simple words,
people’s choices. Recommender system is
algorithm will not be able to differentiate
acting like a friend which helps user to make a
whether user watched action movies first or
decision to get product of their choice from the
horror movies. In this case, the
dearth of options.
recommendations become unreasonable as
Broadly recommender system is divided into
user’s interest has changed. This paper
three main groups known as content-based
proposes a recurrent neural network (RNN)
filtering, collaborative filtering and hybrid
based recommendation model which gives
filtering method. Content-based filtering is
importance to sequence of information’s and
used to provide recommendations to user based
provide appropriate recommendations.
on user’s past history only. Collaborative
The rest of the paper is organized as follows.
filtering is an approach in which similar group
Detailed literature review is discussed in
of people are determined who have same
section 2. Proposed algorithm is explained in
choice as of user for a particular product.
section 3. Data description, evaluation
Hybrid filtering is a combination of both
parameters and experimental results are
contents based and collaborative filtering,
presented in section 4. Concluding remarks are
where user’s past history and common user
given in section 6.
group information are considered for providing
Special Issue on Recent Research Trends in Management, Science and Technology (August 2021) 212
Literature Survey Hinton proposed deep learning-based model on

structured data for providing recommendations
Davidson, James et al. has mentioned how
using Boltzmann machine and neural networks.
YouTube provides their recommendations
Spotify music platform uses deep learning to
using neighbourhood collaborative filtering
analyze the music style and recommend
approach based on user's past history. KNN
according to the style of the song. Google used
considers only last behaviour of user and
deep learning based model for mobile app store
recommends the next item which are more
recommender system which uses cross product
similar to previous one where similarity is
to make embedding vector and give
calculated using cosine similarity. The author
suggestions based on it . Many deep learning
has explained that the shortcomings of K-
based recommendation models uses
nearest neighbour that it scales poorly when
approaches to generate higher order terms
data increases as everyone is becoming
which can handle sparse features, they are
dependent on internet for decision making.
capable of handling wide, deep and cross
Recommender system focus on modelling the
networks. These all networks sum up the
relationship between user and item based on
results from their specialized and Multi-
historical feedback, user's feedback can be
perceptron model passing through a linear
implicit or explicit. Modelling implicit
layer and sigmoidal activation function to
feedback can be challenging due to ambiguity
produce results.
of hidden data. Matrix Factorization (MF)
Despite the success of traditional recommender
methods can be used to uncover latent
system approach several limitations has been
dimensions to represent user -item
identified. Content based recommender system
embedding’s interactions through inner
suffers with popularity biasness. Collaborative
product. Netflix has built recommender system
filtering approach KNN is not able to handle
based on single value decomposition (SVD)
the data sparsity.SVD is able to deal with it but
which has performed better than KNN. SVD++
it does not consider the implicit feedback.
is an extension of regular single value
SVD++ is able to deal with this problem but
decomposition algorithm (SVD) which not
there is a need of deep learning-based approach
only considers the implicit but also explicit
feedback and performs better than SVD[10]. to build recommender system which can
provide some quality recommendations.
Table 1: Analysis of Literature Survey
Existing Methods Scaling Ability Efficiency Sequencing Information
Neighborhood based methods Poor Fair No
Matrix Factorization method Poor Fair No
SVD Poor Fair No
SVD++ Fair Fair No
Boltzmann Machine Good Good No
Neural Network Good Good No
The existing methods for recommender shows the comparison of various methods on
systems are analyzed on three factors these parameters. Based on the existing studies
considered very important in the evaluation. few gaps are identified in the field of
First is efficiency of the algorithm which recommender systems. These gaps are:
explains how many users have actually  The recommender system uses a subset of
considered the suggestions provided by the user experience and provides suggestions.
system. Second is scaling ability. This is a very The input data is not a complete
important factor in upgrading the recommender representation of user experience.
system. If the system is not scalable then it will  Existing system works in a static manner
not allow new data experience to be included in which considers available data and model
the suggestions. Third is the uses of complete the system. Most of the systems are not
user experience in data modelling. Table 1
adaptable to keep adding new data and Recurrent Neural Network layer (RNN):
update the suggestions. Recurrentnt neural network is a type of neural
network where the output from previous step is
ASM: Adaptive Sequential Modelling based
used as input. RNN has a memory through
Recommender System
which they remember the value of input. This
As discussed
ssed in the previous section existing is the base of ASM design as the focus is on
recommender system algorithms do not sequence of user preferences which needs to be
perform well when user’ interest changes over stored for building the recommendation
time. The reason to this is that these algorithms system. Pre-processed
processed data is loaded from the
do not consider user’s changing preference in use case specific dataset to the RNN layer.
account in their decision-making
making process.
proces In this recurrent neural network (RNN) model
Hence, there is a need for new recommendation multiple hidden layers are used. All hidden
technique which considers sequence of layers are combined into a sing single recurrent
interests. The proposed ASM architecture layer, so they are acting as a single unit. Merge
considers the limitations of existing methods, layer is used here to find the dot product of
considers sequence of information and adapt to user and movie vector matrix, to perform final
newly produced data. latent vector matrix. Dropout layer is used here
ASM Architecture
ture to discard all unwanted data to prevent model
from over fitting. Next, fully connected layer is
In this section, complete structure and modules used so that all other layers can get connected
of ASM (Adaptive Sequential Modelling) with each other so that the flow of information
based recommender system is been discussed will be available between all present layers.
in detail. The scope of proposed recommender First the dataset gets loaded and it’s been
system is limited to movie recommendation sorted according to the ttime of rating. For
use case and accordingly data parameters
para are every user u, sequence has been captured based
used for modelling. The broad architecture is on movie id, rating and movie category. Now
divided into three modules: Input layer, RNN this sequence has been prepared for every user
layer and Output layer as shown in figure 1. in the similar manner. After this, this data has
Input layer: Firstly, the data gets loaded from been supplied as a input to first layer of RNN
the dataset and then pre-processing
processing is (Recurrent
current neural network) that is the first
performed to get qualityty data. This step checks layer.
for missing values, duplicate values and noisy After being supplied to the first layer, it’s then
data values. Duplicate values are removed, and applied to RNN layer where it’s been pass
mean value is replaced with missing values. through the different hidden layer for further
Incorrect data are also corrected by replacing preprocessing. Finally output will be getting
with mean values. Two vector matrixes of P, Q the recommendations off movie based on given
are formed based on user and movie. User id, input at the output layer.
movie id is passed as an input which will return Following is the equation to calculate the
the latent factor vectors for the user and movie current state
respectively. ht = f ( ht-1, xt ) (1)
Where ht is current state, ht-1 is previous state
and xt is input state. After calculating the
current state,
tate, next the activation function will
be applied on this as follows
ht = tanh (Whh ht-1 + Wxh xt )
(2) Based on the activation function current
state value will be updated in the back
propagation network.
Figure 1. Model Architecture of Proposed
Algorithm Final output will
ill be calculated as follows
yt = Why ht (3)
Special Issue on Recent Research Trends in Management, Science and

a Technology (August 2021) 214
Result Analysis
In this section various methods of
recommender system are discussed in
comparison with proposed ASM model.
Comparison is done based on error metrics:
Root Mean an Square Error (RMSE) and Mean
Absolute Error (MAE). Both this error is used
to find correctness of predicted result by
comparing it with actual value. In the
following subsequent subsection first datasets
are described in detail then experimental setup
Figure 2. Experimental Setup
iss presented and finally results are discussed in
detail. Here the algorithms are applied on movie lens
(1million) dataset consist of 1 mil million ratings
a) Dataset Description:
from 6000 users on 4000 movies. Experimental
Two different datasets from Kaggle, namely setup is shown in figure 2.
Movie lens 100k and movie lens 1million are a. User Rates movies, next pre pre-processing is
used for data modelling. Movie lens 100k performed to do data cleaning by replacing
dataset has 100,000 ratings (1-5)(1 from 943 all NA values to zero.
users
rs on 1682 movies and each user has rated b. A matrix with user, movies and the
20 movies. It also consists of demographic respective ratings iss created.
information of user related to age, gender, c. These two latent vectors are then supplied
occupation and zip. to the merge layer where the dot product of
Movie lens dataset which has 1 million these two vectors is getting calculated and
information’s with details of user, movie and final vector matrix will be produced.
rating. The dataset hass 1,000,209 ratings of d. Final vector is applied to the dense layer
3,900 movies which has been given by 6,040 which is acting like a fufully connected layer
users. User details consist of user id, age, where each neuron from previous layer is
gender, occupation and zip-code. zip All connected with all other neurons of other
demographic information’s are collected from layers.
the people having age between 1 to 56 years. e. Dropout layer is used here to remove all
Movie details consistist of movie id, title and unwanted data from the network so that
genre. Different types of genres have been used model will not get over fitted which affects
here which includes action, adventure, comedy, the final outcome.
thriller, drama, horror, fantasy, romance, sci-fi,
sci f. Algorithms are applied to make predication
war, documentary. Rating details are made on of most preferable movie for targeted user
the scale of five which is provided by by considering latent features. Based on
individual user for specific movie. target user’s rating for movie is predicted
b) Experimental Setup: by comparing it with another similar user.
g. Here dataset splits the initiinitial dataset into
Mainly train test split procedure is used to training and testing with 60 60-40 criteria.
estimate the performance of algorithm when Next evaluation metrics are used for
they are used to make prediction on data which performance evaluation of algorithms.
is not used to train model. Normally random
splitting has been used for analysis purpose c) Evaluation Measures:
where dataset is randomly divided into samples In this paper, different evaluation measures
but here, we have tested it based on both have been used to evaluate the quality of
random and fixed splitting. proposed recommendation
endation system approach.
a. Root mean square error (RMSE)
Root mean square error (RMSE) is a prediction

error which is used to find the difference Table 3: Comparison of various models of
between predicted and actual value. The lower recommender system based on RSME
rmse value is considered as a best fit and its
Movie
range lies between 0 to 1000. Models Movie lens(1M)
lens(100k)
KNN 0.9912 0.9646
(4) SVD 0.9842 0.9601
SVD++ 0.9723 0.9342
b. Mean absolute error (MAE)
ASM 0.9634 0.9234
It is averaged squared difference between
actual and estimated value .it is basically used Though RSME parameters shows improvement
to capture how accurate model has given result. in ASM for both the datasets but it is necessary
Lower value considered as best fit. to analyze the difference in the values of MAE
and RSME values for both the datasets. Ideally
the difference should not be much.
(5)
d) Performance Evaluation
In this section, various experiments have been
conducted to evaluate the ASM model for
recommender system. In these experiments
ASM is compared with KNN, SVD and
SVD++ based recommender systems. As
explained in the previous section first
experiment
eriment was performed using MAE
evaluation method and as show in table 2. The Figure 3: Difference in MAE and RSME
results show that performance of MAE for 1m values for both the datasets
dataset is approximately 0.04% improved but
for 100k dataset it is approximately 0.02% As shown in figure 3 the difference value of
decreased. So, there was a need to evaluate the
th ASM model for 100k dataset is the lowest as
model on other parameters. compared to another model. Similarly, the
difference value of ASM model for 1M dataset
Table 2: Comparison of various models of is also the lowest. For 100k dataset tthe ASM is
recommender system based on MAE approximately 0.05% and for 1M dataset it is
Movie 0.3% better as compared to other models.
Models Movie lens(1m)
lens(100k)
KNN 0.7823 0.7512
SVD 0.7755 0.7434
SVD++ 0.7615 0.7234
ASM 0.8012 0.7234
Second experiment was performed on another
parameter named as RSME. The method is
same as the previous experiment. All four
models are evaluated based on RSME for both Figure 4: Result analysis on RMSE and MAE
the datasets. As shown in table 3, in this for both the datasets
experiment RSME values have shown In the fourth experiment just to have
improvements in ASM model as compared to confidence in the final conclusion a
other models. Approximately 0.03% for 100k comparison of all thehe models for both the
dataset and 0.05% for 1m dataset. parameters and both the datasets are done. As
shown in figure 4 out of four groups three of
them shows better performance of ASM model.

5: Comparison of KNN, SVD, SVD++ and

ASM
Figure 6. Comparison of loss v/s validation
In the fifth experiment a total comparison of all
loss
the models keeping the oldest model KNN as
the base was performed. As shown in the Based on the above given conditions 18 epochs
waterfall chart in figure 5 SVD shows were performed. As shown in figure 6, ASM
improvement as compared to KNN but not model performed well because there is
better than SVD++ and ASM. SVD++ shows statistically insignificant difference between
improvement when compared with KNN and actual loss and validation loss.
SVD but not better than an ASM. Hence with Conclusion
experiment five it could be concluded that
ASM is better than other models. Traditional recommender system approaches
When the model gets trained it either suffers are most widely used in providing
from an over fitting or under fitting problem. recommendations for a particular product, but
Over fitting is a scenario when the model they faces challenges when user’s preference
performs well on trained data ata but performs gets changed. In this paper, a recurrent neural
poorly on unseen data. This means that model network based deep learning model has been
has memorized the data instead of learning introduced for recommendation system to
relationships between new features. As capture the changing preferences of user.
experiment six model performances are also Proposed algorithm has been tested on two
evaluated based on given conditions:-
conditions: different dataset, based on the observation it
A. if loss > validation loss, then
hen model is under has been stated that proposed algorithm is
fitting performing better with large dataset as compare
B. if loss<validation loss, then model is over to other data set.
fitting Based on experimental result it can be stated
C. if loss == validation loss, means if there is that proposed model is performing better than
very minor difference then model fits perfectly. all other algorithms.
References
1. Davidson, James et al. (2010). “The recommendation systems. In: International

YouTube Video Recommendation World Wide Web Conferences Steering
System”. In: Proceedings of the Fourth Committee, pp. 278––288
ACM Conference on Recommender 3. F. Ricci, L. Rokach, B. Shapira, and P.
Systems. RecSys ’10. event-place:
event Kantor, (2011) Recommender systems
Barcelona, Spain. New York, NY, handbook. Springer US.
USA:ACM, pp. 293–296. 4. Greg Linden, Brent Smith, and Jeremy
2. Elkahky, A.M., Song, Y., He, X. (2015) : York. (2003). Amazon.com
A multi-view
view deep learning approach for recommendations: item
item-to-item
cross domain user modeling in

collaborative filtering. IEEE Internet Guangzhong Sun. xDeepFM (2018)

Computing 7, 1, 76–80. :Combining explicit and implicit feature
5. Guorui Zhou, Na Mou, Ying Fan, Qi Pi, interactions for recommender systems. In
Weijie Bian, Chang Zhou, Xiaoqiang Zhu, Proc. of the 24th ACM SIGKDD
and Kun Gai. (2018) Deep interest International Conference on Knowledge
evolution network for click-through rate Discovery & Data Mining, pages 1754–
prediction. arXiv preprint 1763. ACM.
arXiv:1809.03672,. 12. M. J. Pazzani and D. Billsus, ‘‘Content-
6. Guorui Zhou, Xiaoqiang Zhu, Chenru based recommendation systems,’’ in The
Song, Ying Fan, Han Zhu, Xiao Ma, Adaptive Web. Berlin, Germany: Springer,
Yanghui Yan, Junqi Jin, Han Li,and Kun 2007,pp. 325–341.
Gai.( 2018) Deep interest network for 13. Oord, A., Dieleman, S. (2013): Deep
click-through rate prediction. In Proc. of content-based music recommendation. In:
the 24th ACM SIGKDD International Neural Information Processing Systems
Conference on Knowledge Discovery & Conference, Harra 2013, pp. 2643–2651.
Data Mining, pages 1059–1068. ACM. MIT Press, Massachusetts
7. Hinton, G.E. (2007): Learning multiple 14. Ruoxi Wang, Bin Fu, Gang Fu, and
layers of representation. Trends Cogn. Sci. Mingliang Wang. (2017) Deep & cross
11(11),428–434 network for ad click predictions. In Proc.
8. Hinton, G.E., Osindero, S., Teh, Y.W. ADKDD, page 12.
((2006): A fast learning algorithm for deep 15. S. Rendle, C. Freudenthaler, Z. Gantner,
belief nets. Neural Comput. 18(7), 1527 and L. Schmidt-Thieme, (2009) “BPR:
9. Hinton, G.E., Salakhutdinov, R.R. (2006): bayesian personalized ranking from
Reducing the dimensionality of data with implicit feedback,” in UAI.
neural networks. Science 313(5786), 504– 16. Sarwar, Badrul, George Karypis, Joseph
507 Konstan, and John Riedl (2000a).
10. Huifeng Guo, Ruiming Tang, Yunming “Analysis of Recommendation Algorithms
Ye, Zhenguo Li, and Xiuqiang He. for e-Commerce”, pp. 158– 167.
DeepFM (2017): a factorizationmachine 17. X. Su and T. M. Khoshgoftaar, (2009 )‘‘A
based neural network for CTR prediction. survey of collaborative filtering
arXiv preprint arXiv:1703.04247. techniques,’’ Adv. Artif. Intell., vol. 2009,
I. Portugal, P. Alencar and D. Cowan, (2018) Aug., Art. no.421425.
The use of machine learning algorithms in 18. Y. Hu, Y. Koren, and C. Volinsky, (2008 )
recommender systems:A systematic “Collaborative filtering for implicit
review, Expert Syst.Appl.,vol. 97,pp. 205– feedback datasets,” in ICDM,.
227. 19. Y. Koren and R. Bell, (2011) “Advances in
11. Jianxun Lian, Xiaohuan Zhou, Fuzheng collaborative filtering,” in Recommender
Zhang, Zhongxia Chen, Xing Xie, and Systems Handbook. Springer.
AUTOMATIC SOL-GEL TECHNIQUE BASED ANALYSIS ON OPTICAL PROPERTIES

IN ZNO THIN FILMS
M. Manvizhi1, S. Murugan2 and R. Selvaganapathy3
1,2
A.V.C. College (Autonomous), Mayiladuthurai, Tamil Nadu, India
3
Annamalai University, Chidambaram, Tamil Nadu, India
maanuadhi@gmail.com
_______________________________________________________________________________________
ABSTRACT
In this manuscript, the analysis on optical properties has been carried out through ZnO thin films which were placed
onto the glass substrate employing automatic sol-gel technique, instead of ordinary sol-gel technique. Experimental
analysis has been carried out using ethanol, zinc acetate, diethanolamine and de-ionized water, ZnO films were
organized and produced by fluctuating the number of coatings (2-5 dipping iterations) for zinc acetate concentrations
of 1.0 mol/L. The films were decayed at 300°C for one hour and once the final coating is completed the entire setup has
to be annealed at 5000C for 2 hours. Using UV/Visible spectrometer at 200 to 900 wavelengths, the Optical
transmittance measurements of the deposited films have been carried. Highly transparent visible range can be obtained
fromthe prepared ZnO films. Results found to be suitably declared that the average transmittance of ZnO films for 2
coatings are finally 92% in the visible range with high absorbance in the UV-region, whereas direct bandgap is 3.21
eV and indirect allowed band gap has been found to be 3.08eV respectively.
________________________________________________________________________________
Keywords: Zno thin-films, Zinc Acetate, Automatic sol-gel, Optical properties, UV region, Band gap
Introduction optical gadgets over GaN. Moreover, the

electron versatility of ZnO (115-155 cm2 V-1 s-
Among different functional materials, metal 1
) at room temperature is higher than that of
oxide films are exceptionally appealing for
TiO2 (<10-5 cm 2V-1s-1). Accordingly, with
some applications because of their great
comparative band hole of TiO2, ZnO
electrical and optical properties. Presently a
nanoparticles are in effect generally utilized in
days, zinc oxide (ZnO) has drawn a lot of
the Grätzel-type PV cells. To wrap things up,
consideration of the scientists as one of the
ZnO or In/Al-doped ZnO are as of late being
most encouraging semiconducting materials for
utilized as straightforward conductive oxide
its adaptable applications in blue and bright
(TCO), because of high straightforwardness in
light producer, varistors, substance and gas
apparent reach, as an option of ITO (F:SnO2)
sensors, surface acoustic gadgets, spintronic
and FTO (F:SnO2) on the grounds that they are
gadgets, optoelectronic gadgets and
expensive. Electrical and optical properties of
piezoelectric transducers. Many examination
ZnO thin films are incredibly reliant upon the
bunches are attempting to accomplish p-type
testimony procedure applied. Up until this
conductivity and just as to control inadvertent
point, different affidavit procedures have been
n-type conductivity of ZnO. It is notable that,
utilized to get ready ZnO thin films, for
ZnO takes shape in the wurtzite hexagonal
example, RF magnetron sputtering, molecular
construction with c-pivot direction that
beam epitaxy, spray pyrolysis, hydrothermal
upgrades the anisotropic designs development.
growth, pulse laser deposition, chemical vapor
Yet, the accessibility of ZnO as an enormous
deposition and sol-gel method. However, the
mass single precious stone is the huge benefit
sol-gel method offers brilliant homogeneity,
of ZnO over GaN that likewise takes shape in
decision to utilize both natural and inorganic
the wurtzite structure. Besides, band hole of
antecedents, effortlessness, minimal expense,
ZnO is 3.37 eV at room temperature, while that
great films on enormous region substrate or
of GaN is 3.44 eV. Once more free-exciton
more all, climate well disposed. This work is
restricting energy in ZnO is 60 meV and that of
the continuation of our past work, where ZnO
GaN is 25 mev. Lower band hole and higher
thin films were effectively saved on glass
exciton restricting energy makes ZnO a most
substrate utilizing a custom made length
encouraging material for application in
controlled programmed solgel measure. In past
optoelectronics and excitonic impacts based
work, it was just shown that the planned idle and hanged to the top of the clip. Through
programmed framework can store slender films the motor shaft, the length of the ribbon has to
effectively be that as it may, the pre-arranged be adjusted or varied such that the bottomized
films were not portrayed with any properties. portion of the substrate must touch with the
The current work provides details regarding the precursor solution. Now, the power switch is
examination of optical properties automated turned-on simultaneously any one of the six-
sol-gel inferred ZnO thin films that have been keys has to pressed according to the desired
stored on glass substrate with different number length of the substrate where the thin-films has
of coatings. to be deposited. Immediately, when the motor
starts to rotate the ribbon has to be
Experimental Methodology
disconnected from the motor shaft where the
Fig. 1 inclusively depicts the homemade substrate starts to dip into the precursor
automatic sol-gel system for deposition of thin solution at a constant rate of 1cm/min. As
films. The entire experimental set-up is soon as within 10 secs the motor stops its
organized on a wooden base arrangement. The rotation, when the required length of the
fundamental highlights of the framework are substrate is fully immersed into the solution.
engine, plastic strip, damper weight, cut, At the same rate of dipping 1cm/min, the
measuring glass base and a keyboard. Clip is substrate has to be withdrawn from the solution
connected to a plastic lace like a ribbon another by making the motor again for starting to rotate
finish of which is looped on the engine shaft. A in opposite direction so that the ribbon is coiled
damper weight has been connected with the through the motor shaft. Finally, heat treatment
ribbon to keep it straight and to wipe out the has to be carried out when the substrate is fully
swing and vibration on the grounds that the withdrawn from the solution.
lace is too thin to even consider having nonstop
swing in the normal breeze and presented to A. Chemical Reagents
vibration while the engine pivots. To get a fine ZnO precursor solution was prepared ethanol
wavelength, six keys on the keyboard has to be (M=46.07, CH3CH2OH, EtOH) as so
placed by comparing to six distinct lengths di-hydrate (M=219.50, Zn(CH3COO)2.2H2O,
(2cm, 3cm, 4cm, 5cm, 6cm and 7cm), and a ZAD) precursor, diethanolamine (M=105.14,
power switch. [CH2(OH)CH2] 2 as chelating agent and de-
ionized water.
B. Preparation of ZnO Films
In this work, ZnO films were saved for zinc
acetic acid derivation grouping of 1.0 mol/L
with different number of coatings (2-5
plunging cycles). ZnO forerunner
arrangement was set up in the accompanying
manner. From the outset, 5.488g of zinc
acetic acid derivation was disintegrated in 25
ml of outright ethanol to yield 1.0 mol/L
centralization of forerunner arrangement and
afterward was attractively blended for 30
min. A smooth arrangement was gotten where
with a molar proportion of DEA/ZAD as 1:1,
4.823 ml of DEA and with a molar proportion
of de-ionized water/ZAD as 2:1, 1.8 ml de-
ionized water was added into the emulsion.
Fig.1 Experimental set-up: Automatic Sol-gel
The arrangement was again blended and an
Technique
unmistakable arrangement was acquired after
Once the precursor solution is prepared, the a specific period. In the mean time, a
beaker base and the substrate has to be kept homogeneous blender of total ethanol and de-
ionized water was set up with a volumetric The optical transmittance of ZnO films for 2
proportion of 8:2. At last, the antecedent coatings in the wavelength varies from 300 to
arrangement was hydrolyzed with drop by 900nm. Observance also made that
drop expansion of arranged blender of ethanol electromagnetic spectrum with average
and de-ionized water holding the arrangement transmittance of 92% with 400nm as
under persistent mixing. Glass miniature wavelength as in visible region. A sharp UV
slides were utilized as substrate for testimony also present in 360nm. This transmittance
of ZnO films. The substrates were cleaned value is greater than the earlier mentioned in
ultrasonically, first in ethanol and in this way transmittance value of ZnO films.
in particle traded refined water for 15 min Transmittance extents to its concentrated
each at 500C. After that solvents from the value at greater wavelength.
outside of substrate were cleared away by
utilizing an air blower and kept in a broiler at
800C for 10 min. At that point 2cm length of
the perfect substrate was covered with the
pre-arranged sol utilizing programmed sol-gel
measure. The plunging and pulling out rate
was 1cm/min and stretch between resulting
plunging and pulling out was 10s. The
covered movies were dried at room
temperature for 24 h and afterward warmed at
3000C in air for 1 h to break down the
forerunner film to ZnO. After that another Fig. 2 Images of ZnO films for zinc acetate
testimony cycle was done. In the wake of concentration of 1.0 mol/L with (a) 2, (b) 3, (c)
finishing the necessary affidavit cycles, the 4, and (d) 5 coatings
movies were strengthened at 5000C at a
Fig. 4 shows the retention range of automatic
warming pace of 100C/min and left at 5000C
sol-gel determined ZnO for 2 coatings. It is
for 2 h. The optical properties of the deposited
unmistakably seen that the movies have a low
films were measured by UV/Visible
ingestion at straightforward district and high
spectrometer at 200-900 nm wavelengths
assimilation at bright area. The assimilation
using reference substrate. From transmittance
worth of the pre-arranged ZnO films at bright
spectra absorbance was determined using
locale is lower than recently revealed
Beer-Lambert Law and optical band gap was
information. This is a direct result of the more
found using Tauc relationship.
modest thickness of ZnO films as just 2 coatings
Results and Discussion have been given.
Fig. 2 illustrates the pictorial view of The immediate and aberrant permitted optical
automatic sol-gel produced as ZnO films for advances among valance and conduction groups
different numbers of coatings with zinc can be assessed by fitting a straight line in solid
acetate concentration of 1.0 mol/L. The retention otherworldly area utilizing the Tauc
observance has been made that the all films relationship. As per Tauc law reliance of
are likely to be transparent as visual and the ingestion co-effective (a) on photon energy (hv)
transparency decreases with the increase of can be given by [22]
number of coatings which indicates ZnO films
with 5 coatings reveals greater absorbance (1)
than the films as reduced number of coatings.
Where a is the absorption co-efficient, A is the
The reason for this as, when coating number edge width parameter, and hv is the photon
increases obviously the film thickness also energy, and r is a constant, for direct allowed
has been increases with high absorbance in transition r equals ½ and for indirect allowed
ZnO as routed in when compared with that of transition equals 2.
ordinary sol-gel derived ZnO films.
Fig. 5 Plot of (ahv)2 vs. hv of ZnO thin films.

Fig. 3 Optical transmittance ZnO thin films
Fig. 6 Plot of (ahv)1/2 vs. hv of ZnO thin films.

CONCLUSION
Fig.4 Optical absorbance of ZnO thin films Finally, a good credit of preparation on ZnO
The optical band gap for direct allowed made on glass substrate with the help of eco-
transition of the films have been determined friendly sol-gel technique as a precursor - zinc
from the extrapolation of the linear portion of acetate. A sample of concentration (1.0 mol/L)
(ahv)2 vs. hv at a= 0. Direct transition band has initially been taken to prepare the ZnO thin
gap is found to be of 3.21eV. Although this films trailed with various number of coatings
value is smaller than the bulk value of 3.37 as zinc acetate been a concentrator. ZnO: A
eV but still it is in good agreement with crystal structure has been obtained when
previously reported data of ZnO thin films. deposited films treated with annealing process
Fig. 6 shows the plot of (ahv) 1/2 vs. hv of at 5500C for two hours. A greater transparent
ZnO films for 2 coatings and band gap for at visible region (92%) of ZnO has been
indirect allowed transition has been determined obtained using automatic sol-gel technique.
by extrapolating the straight line portion of the Results reveals that, during direct and indirect
spectrum at a= 0. The indirect allowed band of bandgap with nominal values. Greater
prepared ZnO thin films with 1.0 mol/L and 2 transmittance in the visible region and greater
coatings has been found to be of 3.08 eV which absorption in the UV-region has made the
is greater than the previously reported indirect films potentially able for optical window
allowed band gap of 3.0 eV for ZnO films. applications and also as Transparent
Conductive Oxide (TCO).
References
1. F. E. Ghodsi and H. Absalan, (2010 ) and Y. S. Yu, (2004) “Blueshift of near
“Comparative study of ZnO thin films band edge emission in Mg doped Zno
prepared”, Acta Physica Polonica A, vol. thin films and aging,” Journal of Applied
118, no. 4, pp. 659-664. Physics, Vol. 95, No. 9, pp. 4772-4776,
2. F. K. Shan, B. I. Kim, G. X. Liu, Z. F. Feb.
Liu, J. Y. Sohn, W. J. Lee, B. C. Shin,
3. G. Sberveglieri, S. Groppelli, and G. 12. Muthuvinayagam, N. Melikechi, P. D.

Coccoli, (1988 )“Radio frequency Christy and P. Sagayaraj, (2010)
magnetron sputtering growth and “Investigation on mild condition
characterization of indium-tin oxide preparation and quantum confinement
(ITO) thin films for NO2 gas sensors”, effects in semiconductor nanocrystals of
Sensors and Actuators, vol. 15, no. 3, pp. SnO2”, Phys. B, vol. 405, no. 4, pp.
235–242, Nov. 1067-1070.
4. H. Endo, M. Sugibuchi, K. Takahashi, S. 13. N. Naghavi, A. Rougier, C. Marcel, C.
Goto, S. Sugimura, K. Hane, and Y. Gueary, J. B. Leriche and J.M. Tarascon,
Kashiwaba, (2007 ) “Schottky ultraviolet (2000 ) “Characterization of indium zinc
photodiode using a ZnO hydrothermally oxide thin films prepared by pulsed laser
grown single crystal substrate,” Appl. deposition using a Zn3In2O6 target,”
Phys. Lett., vol. 90, no. 12, pp. Thin Solid Films, vol. 360, no. 1-2, pp.
121906(3pp), March. 233-240, Feb.
5. J. H. Lee and B. O. Park, (2003 ) 14. N. Shakti, P. S. Gupta,( 2010) “Structural
“Transparent conducting ZnO:Al, In and and optical properties of sol-gel prepared
Sn thin films deposited by the sol-gel ZnO thin film”, Applied Physics
method,” Thin Solid Films, vol. 426, no. Research, vol. 2, no. 1, pp. 19-28.
1-2, pp. 94-99, Feb. 15. Ohtomo, M. Kawasaki, T. Koida, K.
6. Janotti and C. G. V. d. Walle, (2009 ) Masubuchi, H. Koinuma, Y. Sakurai, Y.
“Fundamentals of zinc oxide as a Yoshida, T. Yasuda, and Y. Segawa,
semiconductor”, Rep. Prog. Phys., vol. (1998 ) “MgxZn1-xO as a II-VI widegap
72, no. 12, pp. 126501(29pp), July. semiconductor alloy,” Applied Physics
7. K. C. Yung, H. Liem, and H. S. Cho,( Letters, vol.72, no.19, pp. 2466-2468,
2009) “Enhanced redshift of the optical March.
band gap in Sn-doped ZnO free standing 16. S. Ilican, Y. Caglar, M. Caglar and B.
films using the sol–gel method”, J. Phys. Demerci, (2008 ) “Poly-crystalline
D: Appl. Phys., vol. 42, no. 18, pp. Indium-doped ZnO thin films:
185002(5pp), Sept.. preparation and characterization,”
8. L. Xu, X. Lu, Y. Chen, and F. Xu, (2011 Journal of Optoelectronics and Advanced
) “Structural and optical properties of Materials, vol. 10, no. 10, pp. 2592-
ZnO thin films prepared by sol-gel 2598.
method with different thickness,” 17. S. Paul, D. M. F. Hossain, M. H. Islam,
Applied Surface Science, vol. 257, pp. M. F. Ali, and U. Shaha, (2013 )
4031-4037, Feb. “Synthesization of ZnO thin films by
9. M .F. Hossain, Z. H. Zhang and T. length controlled automatic sol- gel
Takahashi,( 2010) “Novel micro-ring process”, Int. Conf. Inf. Elec. Vission,
structured ZnO photoelectrode for dye- vol. 1, pp. 85, May.
sensitized solar cell,” Nano- Micro Lett., 18. S. S. Shariffudin, M. Salina, S. H.
vol.2, no.1, pp. 53-55, Apr.. Herman, and M. Rusop, (2012 ) “Effect
10. M. F. Hossain, S. Biswas, M. Shahjahan, of film thickness on structural, electrical
and T. Takahashi, (2009 ) “Study of sol- and optical properties of sol-gel
gel derived porous ZnO photoelectrode deposited layer-by-layer ZnO
for the application of dye – sensitized nanoparticles,” Transactions on Electrical
solar cells,” J. Vac. Sci. Technol. A, and Electronic Materials, vol. 13, pp.
vol.27, no.4, pp.1047- 1051, Jul/Aug.. 102-105, April.
11. M. F. Hossain, T. Takahashi, S. Biswas, 19. W. R. Saleh, N. M. Saeed, W. A. Twej,
(2009 )“Nanorods and nanolipsticks and M. Alwan, (2012 ) “Synthesis sol-
structured ZnO photoelectrode for dye- gel derived highly transparent ZnO thin
sensitized solar cells”, Electro. Commu., films for optoelectronic applications”,
vol. 11, pp. 1756-1759, July. Advances in Materials Physics and
Chemistry, vol. 2, no. 01, pp. 11-16, Materials. Transactions., vol. 49, no .5,
March. pp. 1186-1191, April.
20. Wacogne, M. P. Roe, A. T. Pattinson, 22. Z. Liu, J. Li, J. Ya, Y. Xin, and Z. Jin,
and C. N. Pannell, (1995 ) “Effective (2008) “Mechanism and characteristics of
piezoelectric activity of zinc oxide films porous ZnO films by sol-gel method with
grown by radio-frequency planar
PEG template”. Materials Letters, vol.62,
magnetron sputtering,” Applied Physics
Letters, vol. 67, no.12 , pp. 1674-1676, no. 8-9, pp. 1190-1193, March.
July. 23. Z. Liu, Z. Jin, W. Li, J. Qiu, (2005)
21. Y. Tsay, M. C. Wang, and S. C. Chiang, “Preparation of ZnO porous thin films by
(2008 ) “Effects of Mg additions on sol–gel method using PEG template”,
microstructure and optical properties of Materials Letters, vol. 59, no. 28, pp.
sol-gel derived ZnO thin films”, 3620-3625, Dec.
CHOICE BASED CREDIT SYSTEM IN INDIA: A CRITICAL REVIEW OF

LITERATURE
F. Lalrinzuali and R.P. Vadhera

Department of Education, Mizoram University, Mizoram, India
lalrinzualifanai@gmail.com
_______________________________________________________________________________________
ABSTRACT
The University Grants Commissions have introduces many plans and policy for the effective functioning of higher
education. As CBCS is a new system there are many challenges such as drafting the entire new syllabus, defining the
credit system, motivating and training teachers, staff and students for smoother implementation, encouraging students’
to stay involved throughout the term and improving attendance, and continuous evaluation. The teachers will also face
many challenges. The academic committees in the universities will have to play an important role to overcome these
challenges and help in preparing the new syllabus in a definite time. This change in the curriculum of higher
education, however, has faced a lot of criticism from educationists, policy makers, teachers and students from across
the country as they find Choice Based Credit System to be impractical in the Indian education scenario. Many of the
higher education institutions of India are still in the pursuit to resist the implementation of the CBCS while others have
welcomed the same, seeing it as an avenue for internationalizing the Indian system of education. The objective of the
following paper is to critically review literature on Choice Based Credit System in India. Based on the analysis of the
review, this paper will look into the benefits, challenges and what are the problems that stand in the way for the
successful implementation of CBCS in higher education.
________________________________________________________________________________
Keywords: Choice Based Credit System, flexible curriculum, open electives, student’s mobility, credit
system.
Introduction Objectives
University Grants Commission has come up 1. To critically review literature on Choice
with the Choice Based Credit System (CBCS) Based Credit System in India.
programme in which the students have a choice 2. To find out the problems faced for the
to choose from the prescribed courses, which successful implementation of Choice Based
are referred as core, elective or minor or soft Credit System in higher education
skill courses, and they can learn at their own institutions of India.
pace and the entire assessment is grade-based 3. To make suggestions for effective
on a credit system. The basic idea is to look implementation of Choice Based Credit
into the needs of the students so as to keep up- system.
to-date with development of higher education Review of Related Studies
in India and abroad. CBCS aims to redefine the A very limited number of researches have been
curriculum keeping pace with the liberalization done in this area as CBCS in the Indian context
and globalization in education. CBCS allows is something new which have been proposed in
students an easy mode of mobility to various the 11th Five Year plan of India for academic
educational institutions spread across the world reforms in higher education, now that the UGC
along with the facility of transfer of credits has made it mandatory to implement CBCS in
earned by students. The CBCS imminently fits all the central universities across the country,
into the emerging socio-economic situations, some articles can be seen online and in recent
and could effectively respond to the journals, but still, not much empirical research
educational and occupational aspirations of the has been taken up on the various provisions
upcoming generations. Aided by modern relating to CBCS, so the scholar could review
communication and information technology, only few empirical researches and theoretical
CBCS has a high probability to be operated articles undertaken between 2012-2021.
efficiently and effectively - elevating students,
institutions and higher education.
Barla (2012) in his study ‘Nuances of Choice teaching hours, maintenance of records, time
Based Credit System for Education and Career spent on evaluation, work load, stress levels
Development’ conducted on a stratified and preparation time. Their Study revealed that
convenience sample of 500 students’ at RG students’ felt that the increased in class size
Kedia Colleges, ST. Mary`s Group of was a major hurdle in the successful
Institutions and Osmania University Campus implementation of the credit system.
Hyderabad, for the purpose of making an Raghvan (2015) in her article ‘Choice based
empirical evaluation of students’` perceptions credit system: A SWOT analysis’ further stated
on CBCS system . The principal Null that the premise behind CBCS is that one
Hypothesis was that students’ do not perceive cannot compartmentalize disciplines/subjects
the CBCS system as a mechanism for in today’s knowledge society and if we provide
enhancement of learning and career such system that students’ will be benefitted
development. Data analysis disproved the the most. It will give the student an opportunity
Hypothesis. It was also found that female to abreast himself/herself about the
students’ understand the CBCS system better developments taking place in the field and
than male students’ and are readily willing for acquire knowledge on it and apply the
a change over from the next academic year knowledge in his/her discipline. Moreover, it
itself. also helps the students’ to develop a taste in
In a study conducted by Roy et al., (2013) on other fields and if possible go for innovation in
‘Attitude towards choice based credit system of his/her interested field. In western countries, it
PG level students’ in higher education: A study is very common that students’ studies inter-
on Assam university’, has taken the sample disciplinary subjects at graduate and post
from PG level Arts and Science students’ of graduate level which help students’ to give
Assam University. The sample size is 56 out of better insight about the relevance of one
which 28 are from Science and 28 from Arts subject in other allied fields. The universities
department. Again, from each discipline equal have the flexibility and freedom in designing
number of boys and girls students’ were the examination and evaluation methods that
selected i.e. 14 boys and 14 girls, the major best fits the curriculum, syllabi and teaching–
findings are: a) Boys are having the highest learning methods, there is a need to devise a
level attitude towards CBCS in comparison to sensible system.
the Girls Students’ of Assam University. b) Hasan & Parvez (2015) extensively discuss
Science Students’ are having the highest level the pros and cons of CBCS. They found the
attitudes towards CBCS in comparison to the percentage based evaluation system as a barrier
Arts Students’ of Assam University. c) There is for students’ mobility, making the higher
no significant difference between Arts and education system rigid. They strongly believe
Science Students’ of Assam University in their that CBCS has makes education learner
attitudes towards CBCS. d) There is no any centered , promotes an interdisciplinary and
significant difference between Boys and Girls intra-disciplinary approach , boosts
Student of Assam University in their attitudes development of professional skills , encourages
towards CBCS. multifaceted development of personality,
Kelkar and Ravishankar (2014) in their study mandates uniformity in evaluation and
‘Choice-based credit system: boon or bane?’ encourages a system of teaching learning. They
concluded that teachers’ being an integral part , however , also realize that implementation of
of the system and valuable stakeholders, it is CBCS suffers from certain issues like –
essential to understand their perspective and increase workload on the teachers , lack of
opinion about the CBCS. Their study analysed infrastructure , maintenance of student records ,
teacher’s opinions, problems faced and compatibility of the main papers , lack of
recommended suggestions for better content mastery among the students and
implementation of the CBCS. In response to subjectivity in evaluation .
the problems faced by teachers’ in the In a study conducted by Chaliha (2015) on
implementation of the credit system, the “Scholastic Achievement of the Post-Graduate
following points we re-evaluated: class size, Students’ of Dibrugarh University: A study of
some Associated Variables.” The differences Aloysius College Mangalore. The study was
between the Post-Graduate students’ of conducted targeting management students’ of
Dibrugarh University studying through Choice- size 150. Findings of their study highlight to
Based Credit System (CBCS) and Non Choice- the fact that education is not the end of process
Based Credit System (Non-CBCS) mode with but an integral part of educational spiral and a
reference to their Scholastic Achievement, well-designed system of evaluation is a
Intelligence, Anxiety, Attitude towards Higher powerful educational device. CBCS is essential
Education, Socio-Economic Status was taken for Higher Education as this system increases
up, the major findings are as follows: it seems the sincerity among the students’ as they
that the students’ studying through Non-CBCS preferred to learn the subjects of their choice.
mode are good achievers than the post-graduate Howlader (2021) the study investigate to find
students’ studying through CBCS mode. The out attitude towards choice based credit system
students’ studying through CBCS mode opined (CBCS) of under-Graduate students in relation
that they have more work load in comparison to their academic achievements. Results reveal
to Non-CBCS mode students’ because internal that there exists significant difference of
assessment marks is increased in CBCS mode attitude towards CBCS between boys and girls
from earlier 25% (Non-CBCS mode) to 40% of students. It is also said that there exist no
the total marks in each course. The study significant relationship between attitude
reveals no significant difference between the towards CBCS and academic achievement of
postgraduate students’ of Dibrugarh University total boys and girls students.
studying through CBCS and Non-CBCS mode
Critical Analysis of the Literature
with reference to their intelligence. It also
reveals no significant difference with Only few serious empirical research has been
reference to their anxiety. It also shows that taken up on the various provisions relating to
they differ significantly in their attitude CBCS, yet when the UGC made it mandatory
towards higher education. As the difference is to implement CBCS in all the central
in favor of the post-graduate students’ studying universities across the country there has been
through Non-CBCS mode, it seems that the many debates, protests and strikes, so many
students’ studying through Non-CBCS mode articles and comments have appeared in the
have more favorable attitude towards higher print and non- print media, many of them being
education than the post-graduate students’ criticisms about CBCS, so the investigator
studying through CBCS mode. As the students’ have highlighted all the relevant related studies
studying through CBCS mode opined that the on CBCS that can be seen so far, this is a clear
courses in CBCS mode are very indication that there is a very limited study
comprehensive and require more time to about CBCS in India. There are many more
complete. The increased mark of internal aspects of CBCS that needs to be identified and
assessment also makes them busier throughout studied more thoroughly. Here are the
the sessions. Over work load, less free time, important areas relating to successful
negative experiences from different dimensions implementation of CBCS.
of higher education etc., creates unfavorable a) Common Minimum Syllabi for the
attitude of the students’ studying through programmes in all parts of the country
CBCS mode towards higher education. It was
found that the students’ studying through Non- One of the salient features of CBCS is that
CBCS mode have better socio-economic status there will be a uniform grading system which
than the post-graduate students’ studying will facilitate seamless mobility of students’
through CBCS mode. across institutions, in India and abroad,
Sumitha, Krishnamurthy & Winfred (2016) regarding this uniform grading system, it is
in their study on ‘An Empirical Study to very difficult for all Universities across the
Measure the Perception of Management country to follow the same pattern, With the
Students’ Towards Choice Based Credit System implementation of CBCS, comes uniformity of
(CBCS): A Case Study’ which was designed in curriculum and evaluation system, this requires
management programme of AIMIT, St. all the institutes of higher education to function
under the UGC. This includes central courses/syllabi/curriculum for its various PG
universities as well as public universities. This programmes under CBCS, has not consulted
further implies loss of autonomy of many the potential employers of its manpower;
universities. Though this looks like a major therefore it will not be easy for them to assess
step towards equalizing education throughout the performance of students’.
the country, this might not go down easily with
e) Cafeteria Approach in Selection of
many universities as sudden loss of power
Courses
means massive changes in the functioning,
administration and management of the The salient feature of CBCS relating to
institutions. This is an issue that needs to be cafeteria approach, wherein students’ will be
looked into seriously, discussed and debated to able to pick courses of their choice has been
find a solution – a way between complete loss partially implemented in many of the
of autonomy and complete bureaucratization. universities in India as: a) Students’ admitted
in a PG Course shall take the choice based
b) Lack of Infrastructure open elective courses offered by the
Lack of infrastructure seems to be a huge Departments/Faculties, Sometimes due to lack
problem in almost all higher institutions in of infrastructure and limitation in number of
India. The most important aspect of CBCS, faculty the students’ cannot opt for the OE
which is Open electives, cannot be courses of their first choice.
implemented fully, and student’s mobility is
f) Investment of Time in Learning, Not In
not possible. The students could not get their
Teaching
first choice and in some case not their second
or third choice of open electives because of One of the objectives of CBCS is based on the
lack of seats in that department. So, proper investment of time in learning, not in teaching,
planning and designing on the part of the but in many Universities is given more
stakeholders is required in this matter. importance even in the classrooms. Under the
CBCS, adopted many Universities, the time
c) Increase in work Load of the Teacher spent on Lectures (L), Tutorials (T) and
There is a general perception in the university Practical/Practicums (P) is reflected separately
that with the introduction of CBCS the under L, T and P in the Course Structure of
workload of academic departments has every discipline. Continuous and
increased to a great extent, as it demands too comprehensive evaluation of students’; which
much involvement of the department in testing includes class tests, assignments, and seminars
and evaluation, compiling of results and other etc.; is done under C1 and C2 that relate to the
clerical work. The teachers, under CBCS will first half and second half of the semester,
have to be continuously involved in preparing respectively. And C3 relates to the end
lectures for the different papers, maintain semester examination carrying 60 per cent
records of the students and evaluate and assess weightage, the time taken for learning and
students, their papers, their projects and their preparation for examination are not taken into
continuous growth. Therefore, with the consideration.
implementation of Continuous and
g) Self-paced Learning
Comprehensive evaluation (CCE), under
CBCS, the teachers will be continuously under The CBCS adopted by the Mizoram University
excessive pressure, which will leave them with does not support the self-paced learning as it is
no time for personal professional development. not much different from its traditional system
(Non-CBCS) in terms of its provisions on self-
d) Consultations with Potential Employers paced learning. All students’ of a particular
One of the features of CBCS is that it will help department or school are required to undertake
potential employers, and will make the process the same number of credits. The provision to
easier when they have to assess the not repeat all the courses in a given semester if
performance of students’. Many Universities in a student fail in one or more courses was also
the country while planning and developing there in the non-CBCS.
h) Transferring of Credits 3. CBCS puts the focus on the 'learner', thus

there is a need to give students’ the
CBCS offers the opportunity to study at
flexibility to choose a mix of courses and
different times and in different places. Credits
encouraging interdisciplinary studies. To
earned at one institution can be transferred to
resolve this problem the intake capacity of
another. Unfortunately, none of the universities
the academic departments for Open
in the country has yet implemented this salient
Electives should be increased so as to meet
feature of the CBCS. As of now the students’
the demands of the students’.
are supposed to earn all their credits from their
4. CBCS allows ‘seamless mobility' across
parental institution. Absence of this the
higher education institutions and transfer
provision of transfer of credits from/to other
of credit earned by students’. Yet there is
institution has killed the basic spirit or
no clarity on how students’ movement will
philosophy of CBCS. In some universities
take place. So until and unless all the
where MOOCs have been introduced the
universities and colleges follow the same
students will be able to earn extra credits.
grading system and syllabus this is not
i) Amalgamation of Skills Development possible. Universities all over the country
Courses in the Curricula should aim at uniform grading system
Another important objective of CBCS is which will facilitate seamless
amalgamation of Skills Development Courses mobility of students’ across institutions in
in the Curricula. Since most of the FCs, CCs, India and abroad.
SCs and OEs offered by various departments 5. As CBCS focuses on the employment of
are theoretical in nature, therefore, it can be students’, open electives offered by
said that the CBCS adopted by Universities has academic departments should be skill
not focused on the amalgamation of skills oriented.
development courses in the Curricula. It is such 6. According to UGC guidelines, in the new
that it will not be able to meet the needs of system there is a shift in focus from
national and international employment market teacher-centric to learner-centric
which requires individuals with extraordinary education. Emphasis is on learning and not
skills, vast knowledge and strong moral values; on teaching, with the learner being at the
as such components have not been centre stage of all academic transactions.
incorporated in it. So the university has to change its system
so that the students’ can move in their pace
Suggestions and style.
In the light of the analysis of the review, the 7. CBCS encourages teaching through
researcher makes the following suggestions: projects and case studies. By enabling
students’ to access and critically review
1. As CBCS has many new unique features, research papers and projects online, and
orientation programmes and workshops effectively use them for self-study,
should be organized for teachers’, students teachers’ will end up providing students’
and academic staff to understand the with critical thinking skills. This requires
details about grading, semester, credit moving away from the traditional
system, credit transfer etc. by inviting and classroom and textbook based approach
involving resource persons with but our universities do not have adequate
appropriate expertise. infrastructure to facilitate this, including
2. Another issue is that there is a cafeteria of smart boards, projectors, access to
choices in CBCS but the question is on computers and Internet. So, measures have
how one knows which combination works to be taken to have the required
best, so there is a need for Academic infrastructure for effective implementation
Adviser in each department who will assist of CBCS.
the students’ in their selection of courses
and they should play their role in assisting
the students’ in selection of their courses.
Conclusion present scenario of different universities across

the country, the differences between Central
The University Grants Commission have come
University, State University and Colleges in
up with many new policies and changes in
terms of efficient teachers, academic
higher education which aims to bring
environment, infrastructure etc., following
efficiency and excellence in the higher
uniform grading system and maintaining
education system in India, right from the
harmony is going to be a huge task which will
curriculum to the teaching-learning process, to
require lot of handwork on the part of all
examination and evaluation systems. Looking
stakeholders and close watch have to be kept
at the objectives of Choice Based Credit
for effective implementation of CBCS.
System, it can also be regarded as the right step
for quality education. However, looking at the
Refereces
1. Barla, V. (2013). Nuances of Choice Based 5. Hasan, M. & Parvez, M. (2015). Choice-
Credit System for Education and Career Based Credit System in India: Pros and
Development. Retrieved October 18, 2016, Cons. Department of Education, Aligarh
from Muslim University, Aligarh- UP. India.
https://eltweekly.com/2013/05/nuances-of- Journal of Education and Practice. 6(25),
choice-based-credit-system-for-education- 2015. Retrieved on 17th. April. 2017 from
and-career-development-by-dr-v-barla/ www.iiste.org
2. Chaliha, A. (2015). Scholastic 6. Howlader T (2021) Attitude towards
Achievement of the Post-Graduate Choice Based Credit System (CBCS) of
Students’ of Dibrugarh University: A study under- Graduate student’s in Relation to
of some Associated Variables. Accessed their Academic Achievements.
and Retrieved on 6th. July. 2017 from https://www.ijcrt.org/papers/IJCRT210208
http://shodhganga.inflibnet.ac.in/bitstream/ 6.pdf
10603/32764/12/12_chapter%206.pdf 7. Kelkar, A. S., & Ravishankar, L., (2014).
3. Chaubey, A. K., & Vishwavidyalaya, H. Choice-based credit system: boon or bane?.
G. (2015). Choice Based Credit System Current Science. Guest Editorial. 107(8).
(CBCS): A Better Choice in Education 8. Roy, N. R., Khanam., U. K., & Devi, T.
System. International Journal of Creative (2013). Attitude towards choice based
Research Thoughts. 3(6), June 2015. credit system of PG level students’ in
Retrieved on 1st Nov. 2016. from higher education: A study on Assam
http://ijcrt.net/UploadedArticle/176.pdf university. Scholarly Research Journal for
Interdisciplinary studies, I (1), 1198-1208.
4. Deuri, C. (2015). Attitude towards Choice
9. Sumitha, P.V., Krishnamurthy, M. G., &
Based Credit System of Post Graduate
Winfred, B. R. (2016). An Empirical
Level Students’ in Higher Education: A
Study to Measure The Perception Of
Study on Guwahati University.
Management Students’ Towards Choice
International Journal of Interdisciplinary
Based Credit System (CBCS): A Case
Research in Science Society and Culture
Study. IOSR Journal of Business and
(IJIRSSC). 1(2), (December Issue), 2015.
Management (IOSR-JBM), 56-65.
Retrieved on 16th. April. 2016. from
10. U.G.C. (2008). Guidelines on Adaptation
http://www.academia.edu/20119082/Attitu
of Choice Based Credit System. University
de_towards_Choice_Based_Credit_System
_of_Post_Graduate_Level_Students’_in_Highe Grants Commission. Bahadurshah Zafar
r_Education_a_Study_on_Gauhati_University Marg New Delhi. 110 002.
DECEPTIVE SUBLIMINAL STRATEGIES IN ADVERTISING: A SEMIOLOGICAL

STUDY ON THE SUBLIMINAL STRATEGIES IN ADVERTISEMENTS OF
MCDONALD’S VALENTINE’S DAY PRINT CAMPAIGN.
Gopakumar AV1, Renjini T2 and Sudheer S S3

1
Department of Media Studies, Kristu Jayanti College, Bengaluru, India
2
Department of Psychology, Govt. College for Women, Thiruvananthapuram, India
3
Department of Film Studies, Thunchath Ezhuthachan Malayalam University
_______________________________________________________________________________________
ABSTRACT
The paper focuses on subliminal strategies used in McDonald’s Valentine’s Day print campaign created by TBWA -
Thailand. The Semiological review of the print advertisements revealed that the advertisers have adroitly used
subliminal strategies of Dr.Wilson Bryan Key’s -Figure-Ground Reversal, Embedding, Double entendre, and Low-
intensity lighting. Furthermore, researchers also could find appropriate viewing patterns, design principles like
emphasis, techniques like texture segregation, eye gaze, cross-cultural references, and subjective shots to enhance the
qualities of subliminal messages in the campaign. These creative strategies of the advertisers can tap into the
vulnerabilities surrounding the subconscious and unconscious minds of the viewers, manipulating and controlling them
to bring in persuasive buying behavior.
________________________________________________________________________________
Keywords: Subliminal Advertising, Figure-Ground Reversal, Embedding, Double entendre, Low-intensity
lighting
Introduction example, audio stimuli may be played below

normal human audible volume range or
Today we live in a highly consumerist society
masked by another audio stimulus. Advertisers
where the consumer is king. There is intense
use these techniques to manipulate and control
competition among companies to reach out to
us in many ways. Despite being unethical it is
consumers and persuade them to buy products.
often difficult to prove the use of subliminal
It is here that advertising plays a major role.
advertising. Therefore authorities often become
Advertisers use every method available to draw
helpless in bringing legislation to curb it. It
attention to the product, convince and persuade
has been demonstrated through studies that
people to buy the product. Often many
subliminal messages can stimulate mild
techniques deployed in the field of advertising
emotional activity but these messages can
are borrowed from concepts in Psychology like
trigger action subliminally only if the person
attention, perception, illusion, persuasion, etc.
intends to perform the specific action.
The one technique often used but which
The word “subliminal advertising “was first
advertisers do not acknowledge is that of
coined by James Vicary in 1957. He claimed in
subliminal perception,i.e., perception below
a study on forty-five thousand moviegoers that
the threshold level. This is in contrast to
when the message ‘eat popcorn drink Coca-
supraliminal stimuli or perception above the
Cola” was flashed briefly on the screen, sales
threshold.
increased by 57%. Though controversies are
Functional MRI studies have shown that
surrounding the authenticity of the study, new
subliminal stimuli trigger definite regions of
studies in this area claim that subliminal
the brain regardless of participants being
advertising is a reality. A study by psychologist
unaware (Brooks et al., 2012). Visual stimuli
Hawkins has proved increased thirst ratings
may be rapidly flashed before an individual can
following subliminal exposure to Coco-Cola
perceive them or flashed and then masked,
advertisements and that subliminal advertising
thereby disturbing the processing of the brain
(Krosnick et al., 1992). The efficiency of can affect consumer behaviour. Alison
Motluck of New Scientist showed in a study
subliminal messaging has been authenticated to
how when thirsty participants who were
prime individual responses and stimulate mild
exposed to subliminal advertising of Lipton Ice
emotional activity (Bar et al., 1998). For
were prone to choose the drink afterward.
Follow-up studies of Hawkins could not find advancements in graphics and technology,
convincing results. There have been arguments several tools are available with advertisers to
against the concept of subliminal advertising expertly hide subliminal messages. Through
and studies have tried to prove that subliminal this study, researchers try to explore the
advertising does not exist. Rebecca Clay‘s subliminal strategies in advertisements of
article in American Psychological Association McDonald’s Valentine’s Day print campaign
gives cases of psychologists like Esther created by TBWA-Thailand.
Thorson, who are giving insights about this art
Methodology
of persuasion and the strategies behind the
making of subliminal messages. It is also The research method followed is a qualitative
believed that subliminal messages are most study, using techniques of content analysis and
likely to affect individuals who believe that Wilson Bryan Key’s subliminal strategies with
such messages will alter their behaviour. An semiology. Content analysis is typically called
action can be triggered subliminally only if the quantitative as it involves counting and
person has a predisposed mindset to perform summing phenomena, to even support studies
this specific action (Key, 1976). of more qualitative nature. Krippendorff (1980)
Vance Packard has distinguished calls it primarily a symbolic method as it is
consciousness into three levels namely used to investigate symbolic material,
conscious –rational, preconscious, or conducting many interpretative works, relying
subconscious, and unconscious. In the on a good knowledge of the texts under
conscious rational level, one is aware of what examination. This method of textual analysis is
is going on and is even able to tell why. At the frequently combined with other methods in
preconscious level, the person has a vague qualitative research. Here, a combination of
awareness of what is going on with his feelings semiotic analysis, syntagmatic as well as
attitudes, etc. but is not able to tell why. It is at paradigmatic analysis of the print
this level one’s fears, desires prejudices, etc. advertisements is used considering them as
operate. The third level or unconscious is discourses of a continuous nature. Furthermore,
where we are not aware of our feelings which to study the three print advertisements as a part
are repressed and we cannot discuss them. The of McDonald’s Valentine’s Day print
target of advertisers, especially in subliminal campaign created by TBWA-Thailand,
advertising is the second and third levels of researchers have specifically used Wilson
consciousness. In other words, at the conscious Bryan Key’s -Figure-Ground Reversal,
level or the visible level, a safe, neutral, or Embedding, Double entendre, Low-intensity
naturally appealing ad is presented, but they are key lighting, and other subliminal techniques in
targeting the subject’s desires fears, etc. at the the advertisements (Key, 1980). The
subconscious or even unconscious level (Key, advertisements were created by TBWA
1973). Thailand, published in February 2020, and
According to Wilson Bryan Key, subliminal posted to McDonald's Thailand's social media
advertising is a double-edged sword. At the channels to mark Valentine’s Day
conscious level, the message does not create Figure-Ground Reversal in McDonald’s
anxiety and prevents perceptual overload. At Valentine’s Day Print Campaign
the same time, the processing is taking place at
the subconscious level and unconscious level When three advertisements were analyzed with
and these events influence behavior (Key, semiotics, researchers could see that campaign
1973). created by TBWA – Thailand brings romance
When subliminal messages use images of sex to the campaign and uses burgers as a sign to
or fear, the defense mechanisms block the signify the facial parts (lips and nose) of an
explicit message but the unconscious mind individual. Every perceived image has a figure
perceives it. Therefore processing still happens (foreground, subject) and ground (background,
in the unconscious mind with the message and environment), which is differentiated by our
product and later the sight of the product at the perceptual defense mechanism to protect the
store may lead him to buy it. With the brain from sensory overload (Key, 1980).
Therefore, viewers consciously notice the Furthermore, their hidden meanings have
figure whereas the ground undulates around it sexual connotations. For example, the
unless it’s not triggered to the foreground McDonald’s slogan “I’m loving it," when taken
through some design elements. Perceptual out of context, has sexual implications.
psychologist Dr. E. Rubin created his famous According to Key, double meanings appear to
Rubin's Profiles to explain this concept. His strengthen the significance in virtually any
profiles, the faces and vases, old women and symbolic stimuli. This unconsciously perceived
young women, duck, and rabbit, are information of a socially objective nature
syncretistic (two-sided) illusions (Rathus, ensures a deep emotional response and
2011). Viewers who notice only one set of continued memory (Key, 1980). Moreover, the
features will miss the other and construe slogan is conveniently placed within the
different meanings, and advertisers take sexually instigating visual to trigger the
advantage of this to create subliminal messages subconscious mind of the viewer.
into the picture's background. These images The three advertisements suggest that the
created as subliminal are usually cultural experience of eating the product is somehow
taboos, making it even tougher for the viewers equivalent to sex. This is the metaphorical part
to perceive it of the food-sex connection, maybe because
In this ad campaign, one can see a man/woman both food and sex are part of the basic needs
attempting to engage in a lip lock kiss in the namely the physiological needs in Abraham
background and in the foreground man/woman Maslow’s hierarchy of needs (Mangal, 1990)
trying to bite a burger, making it a convincing A psychoanalytically inclined observer might
example of Figure-Ground Reversal for believe that this kind of association is also a
embedding subliminal message. feature of some people's actual experiences of
Furthermore, advertisers have used textual sexuality. (Patterson & Kim, 1991). Another
segregation to differentiate the foreground and aspect noticed by the researchers is the
background. In images, texture can be the positioning of this slogan in the lower half of
optical grain of a surface, and one the advertisements, that too close to the bottom
unconsciously unifies objects into regions that borderline. Research shows that objects in the
are confined by an abrupt change in texture. top half of a picture are considered to be more
Here, in the three advertisements, researchers active, dynamic, and potent than what is placed
have noticed a smooth colored background and in the bottom half (Niekamp, 1981). Another
a grainy foreground. Moreover, through texture study found that viewers spend more time
segregation, the foreground can be separated viewing areas appearing on the left and upper
from the background. The association between half of the field than on areas located on the
figure and ground is a prerequisite for right and lower half (Goldsmith, 1984).
perceiving shapes and eventually identifying In this campaign, two out of three
objects. In addition to this, texture perception advertisements, the slogan is placed in the
also presents spatial information by providing lower right portion and one on the lower right
cues for depth perception (Gibson, 1986). portion of the ad copies, and this varying the
Considering these design intricacies, graphic position of the slogan in a frame changes its
designers manipulate these individual impact on the observer by creating signs for
properties to convey meaning. In the three conscious, subconscious and unconscious
advertisements chosen for study, texture levels but predominantly for subconscious and
simulates surface qualities to help us identify unconscious levels for signifying the
and recognize the human faces in the ad copy. subliminal message.
Double Entendre in McDonald’s Valentine’s Embedded technique in McDonald’s
Day Print Campaign Valentine’s Day Print Campaign
Many images, phrases, and slogans have Embedding is a complex process of masking
concealed double meanings even when it one image in the form of another in subliminal
appears normal. This is called double entendre. advertising. According to Key, these embedded
Symbolism also comes under this category. components improve the perceptual experience
of the images, and this suppressed information Low-intensity

intensity lighting Technique in
stays in our memory system for a longer McDonald’s Valentine’s Day Print
duration, evenn for a lifetime (Key, 1980). Campaign
Private body parts are widely used for It’s a subliminal strategy and refers to any
embedding, and graphic artists with expertise image with a predominance of shadowy areas.
bring in these images in an implicit manner to In Subliminal advertising, graphic artists make
cater subliminally. In this Campaign, the logo indistinct subliminal images below the
‘M’ resembles female breasts, and the shape of conscious threshold of perception but beyond
the letter ‘M’ subconsciously arouses appetite the unconscious perception threshold (Key,
and conforms to the oral stage of Sigmund 1980). In this campaign, the advertisers
Freud’s psychosexual stages. This logo was intentionally underplay the image of a burger
made with the suggestion of psychologist by masking with silhouettes and redirects the
Louis Cheskin in 1962. sexual image hidden as a figure-ground
reversal to the subconscious mind to trigger
their buying desires.
Furthermore, low-intensity
intensity lighting is creating
the silhouette of a burger in the three
advertisements, and they typically depict a
form through the outline of its shape, an
interior without detail or texture, and a fill of
uniform color often within a flat pictorial
space. In this campaign, burger silhouettes
promote quick perception and speedy
comprehension of human faces. Graphical
artists’ creative play offers many possib
possibilities
in silhouette for expression, and can quickly
According to Freud, personality develops in convey a sense of mystery, representing a
five psychosexual
exual stages, and of these, the first shadowy world devoid of detail. A well well-
three stages happen before puberty. The first designed silhouette is a solid expression of
stage is the oral stage i.e. from birth to about compacted information that can be understood
eighteen months. At this stage, the mouth is the with a minimum of cognitive eff effort (Malamed,
pleasure zone where the infant gets pleasure in 2009). To avoid perceptual ambiguity,
feeding and the mouth is the he erogenous zone. advertisers have given a formless region for the
Fixation occurs if the needs at any of these background and differentiate the figure with a
stages are under met or overly met. In the oral well-defined
defined boundary or color contrast.
stage, fixation happens when the child is Besides these, advertisers have decreased the
dissuaded too early or late i.e. even after size of the Burger as compared to the explicit
infancy. Freud believed that many of the human faces in these advertisements, for the
personality manifestations
estations can be traced back reason that smaller shapes tend to be perceived
to childhood and if the fixation happens at the as the figure rather than the ground.
oral stage the person goes on to develop habits This creative effort of the advertising company
with pleasure-seeking
seeking through the mouth. He also allows the viewers to perceive sexuality in
believed that smoking, alcohol abuse, nail- nail the
he form of a lip lock kissing in the image with
biting, and overeating can be explained in this
t top-down
down processing, where visual awareness
manner (Ellis et al., 2009). Subliminal female is driven by the viewer’s memories,
breasts in the logo target the subconscious and expectations, and intentions.
unconscious minds of adults who are
experiencing fixation in the later stage. Viewer Navigation in Advertisement and its
Moreover, this sign also elicited appetite in Emphasis on Subliminal Elements.
such people and leads to junkie food habits. The three selected advertisements
ertisements taken for the
study are following the Z Z-pattern layout, and
the eye route tracks the letter ‘Z’. That means focal point and signifies intense passion with
viewers will start in the top-left region, then closed eyes, opened mouth, and a leaning face.
move horizontally to the top-right portion of Another connecting aspect with this is the ‘eye
the image and then diagonally to the bottom- gaze’, which draws viewers to people, as
right portion of the image before finishing with human brains have a specialized mechanism
another horizontal movement to the bottom- for recognizing human faces. Irrespective of
right portion (Lidwell et.al, 2010) The whether the face appears as a photograph, a
advertisers have chosen this pattern as it is painting, a sketch, or a simple schematic
good for simple designs with a few key figure, definite neural networks are triggered in
elements that require to be seen. In the three the brain upon perceiving anything configured
advertisements selected for study, there are as a human face (Malamed, 2009). Another
only a few key elements like a human face, intriguing characteristic of facial awareness is
burger, logo, and slogan, and this pattern easily that viewers automatically shift their eyes in
persuades the viewer to go in that direction to the direction where the face in the image is
help them perceive the intended message of the looking, and many studies have proved that the
advertisers. According to Caleb Gattengo, a eye gaze is used to orient a viewer’s attention
graphic artist will lose control if the viewers (Jason, 2002).In this campaign, advertisers
are permitted to wander at will through the craftily guide the viewers to the burger which
work (Malamed, 2009). is subtly imaged like another face through the
dominant human faces in the advertisements
and it signifies a passionate lip kissing act.
In addition to this, human faces in the
advertisement campaign are captured through
subjective shots and this creates an image's
simulation of someone's visual experience.
Subjective shots can be taken as critical
examples of iconicity, and the iconicity of
Furthermore, this ‘Z’ pattern layout connects visual images serves this process by making it
all the major elements in a hierarchical order possible for images to draw associated
by keeping the logo and human faces on the emotions to which we are already attuned
prominent space, i.e. the upper half and the through our interactions with our social and
other elements like burger and slogan in the natural environments. Additionally, the explicit
comparatively less significant lower half, human faces used in the campaign have cross-
giving space for the subconscious and cultural references as the advertisers have used
unconscious mind to draw the subliminal the images of Caucasian and Mongoloid faces
message. to have a global effect (Messaris, 1997).
Directing the eyes serves two principal needs, Kernan and Domzal (1993) in an article
one to steer the viewer’s attention along a track entitled "To globalize, visualize," opines that
according to the planned ranking order and to the powerful global advertisements are never
capture the viewer’s attention to precise verbose as verbal advertisement requires the
elements of importance. With the design consumer to understand the complexities of
principle “emphasis’, the designer can give language in which it is communicated, whereas
relative dominance to one element and can as visually predominant advertisements give
direct the eyes around that graphic. In this flexibility to the consumer to interpret the
campaign, explicit human faces are given super advertisement theme in an individualistic
dominance through their bigger size. Rudolf manner because of its universality. Here in this
Arnheim in his book Art and Visual Perception campaign, facial expressions of human faces in
notes that when all other factors are equal, the the advertisement copies, gestures, postures,
visual weight of an element is most dependent cultural references, are capable of simulating a
on its size (Malamed, 2009). Moreover, this sexual experience in the viewers who have
super dominant element becomes the primary experienced or repressed this kind of feeling in
their memory. These expressions, gestures,
postures act as subliminal stimuli to evoke the and interpretations of a picture. In this study,
feelings of sexuality in the subconscious and researchers could find appropriate viewing
unconscious. patterns, design principles like emphasis,
Conclusion techniques like texture segregation, eye gaze,
cross-cultural references, and subjective shots
Caleb Gattengo said the vision is quick,
to enhance the qualities of subliminal messages
complete, analytical, and comprehensive. It
in the campaign.
uses very little energy to complete at the speed
of light. It allows human thought to receive and Besides these, researchers found prevalent
store an unlimited amount of information usage of subliminal advertising strategies in
elements in a fraction of a second. (Malamed, McDonald’s Valentine’s Day print campaign,
2009). This design concept works effectively in when they analyzed the three advertisements of
subliminal advertising. Subconscious elements campaigns, they found Wilson Bryan Key‘s
are misleading in nature because consumers subliminal strategies like -Figure-Ground
will not perceive them at a normal level of Reversal, Embedding, Double entendre, Low-
consciousness and therefore cannot control intensity key lighting in the campaign selected
their acceptance or rejection of information. for the study. The most interesting fact is that
Unlike normal advertising, subliminal
all these subliminal techniques are operating at
advertising cannot be curbed by legislation as
it's not operating at the conscious level. the subconscious and unconscious levels.
Advertisers take benefit of the susceptibilities These operating platforms will allow
in our subconscious and unconscious minds, advertisers to reject any such strategy and
using graphical imaging, and they have claim that if you look long enough, you will
immense resources to manipulate each image find it. But the most important evidence is that
to target a specific weakness in people. these advertisers spend billions of dollars each
Furthermore, a picture is more than a two- year on research and subliminal advertising.
dimensional surface, it reflects the designer’s Therefore, the way out here is that advertisers
intent and signifies there is information to be should be ethically accountable to the
communicated. The created graphics is a result
consumer, and from the consumer side, they
of creative play and thoughtful decisions,
evoking a visual experience. Above all, should not buy a product merely based on
designers decide how to see a graphic in an advertising and packaging.
orderly sequence and alter viewer’s perceptions
References
1. Bar, M.; Biederman, I. (1998). "Subliminal 5. Goldsmith, Evelyn (1984) Research into
Visual Priming". Psychological Science 9 Illustration: An Approach and a Review.
(6): 464–469. Cambridge: Cambridge University Press.
2. Brooks, S.J.; Savov V, Allzén E, Benedict 6. Kernan, J. K., & Domzal, T. J. (1993).
C, Fredriksson R, and Schiöth HB. (2012). International advertising: To globalize,
"Exposure to subliminal arousing stimuli visualize. Journal of International
induces robust activation in the amygdala, Consumer Marketing, 5(4), 51-71.
hippocampus, anterior cingulate, insular 7. Key, B. W. (1973). Subliminal seduction.
cortex, and primary visual cortex: a NJ: Prentice-Hall.
systematic meta-analysis of fMRI studies.” 8. Key, B. W. (1976). Media sexploitation. N
NeuroImage 59 (3): 2962–2973. J: Prentice-Hall.
3. Ellis, A, Abrams, M & Abrams, L. (2009) 9. Key, B. W. (1980). The clam-plate orgy
Personality Theories: Critical Perspectives, and other subliminal techniques for
London: Sage Publications. manipulating your behavior. N J: Prentice-
4. Gibson, James J. (1986). The Ecological Hall.
Approach to Visual Perception (Mahwah, 10. Krippendorff, K. (1981). Content
NJ: Lawrence Erlbaum, 1986. Analysis; An Introduction to its
Methodology, Beverly Hills, CA: Sage 15. Messaris, Paul (1997). Visual Persuasion –
Publication The Role of Images in Advertising,
11. Krosnick, J. A.; Betz, A. L.; Jussim, L. J.; London: Sage Publications.
Lynn, A. R. (1992). "Subliminal 16. Niekamp, Walter (1981) “An Exploratory
Conditioning of Attitudes". Personality and Investigation into Factors Affecting Visual
Social Psychology Bulletin 18 (2): 152 Balance,” Educational Communications
12. Lidwell, W., Holden, K. & Butler, J. and Technology 29 (1981): 37–48.
(2010) Universal Principles of Design, (Niekamp, 1981)
Revised and Updated: 125 Ways to 17. Patterson, J., & Kim, P. (1991). The day
Enhance Usability, Influence Perception, America told the truth. New York: Plume.
Increase Appeal, Make Better Design 18. Rathus, L. F. (2011) Foundations of Art
Decisions. Beverly: Rockport Publishers and Design: An Enhanced Media Edition
13. Malamed, Connie (2009) Visual Language San Francisco, CA: Cengage Learning
for Designers: Principle for creating 19. Tipples, Jason (2002)“Eye Gaze Is Not
Graphics that people understand, Beverly: Unique: Automatic Orienting in Response
Rockport Publishers. to Uninformative Arrows,” Psychonomic
14. Mangal, S.K. (1990) General Psychology. Bulletin & Review 9 (2002): 314–318.
New Delhi: Sterling Publishers Pvt. Ltd. (Jason, 2002)
DEEP LEARNING FOR SENTIMENT ANALYSIS: A REVIEW
A. Save* and N. Shekokar

Computer Engineering Department, University of Mumbai, India
*ashwini.save@gmail.com
_______________________________________________________________________________________
ABSTRACT
The intriguing challenges of performing sentiment analysis has kept the researchers on their toes. Because of this
sentiment analysis has been a very active area of research in the present times. And, the other reason why analyzing
user sentiments has been an active area of research in recent times is the varied applications it has in the corporate
sector. In fact, it would not be wrong to say that many businesses have flourished and prospered only on analysis of
sentiments. Consequently, many techniques have been experimented for undertaking effective sentiment analysis.
Techniques such as Natural Language Processing, and Machine Learning techniques have been employed for
performing intuitive sentiment analysis. The recent successes of Deep Learning techniques in many different domains
have mandated its experimentations in the sentiment analysis domain as well. Even in the sentiment classification
tasks, deep learning techniques have shown promising results. The paper performs a detailed and critical review of
recent experimentations carried out for sentiment classification using various deep learning techniques..
________________________________________________________________________________
Keywords: Sentiment Analysis, Deep Learning, RNN, CNN and RCNN
Introduction they have bought.

So, it becomes extremely important for these
The number of people logging onto the
companies to scan these reviews and extract
internet has been growing at an unprecedented
actionable intelligence from these reviews
rate, especially in the past decade. Also, with
from the customers. Extracting relevant
the advent of smart phone technology and
information from the thousands and thousands
incremental growth in new network
of reviews which are written daily manually is
technology have played a very crucial role in
tedious task and prohibitively expensive.
lending a helping hand in this exponential
Hence, a method or technique is required
growth in the internet user base.
which can automatically scan through these
With the rapid innovation in new networking
reviews and extract actionable information.
technologies the cost of accessing and using
One such method which can be used for
the internet has reduced considerably. On the
extracting relevant information is the field of
other side the internet speed has gone up
Data mining, especially sentiment analysis.
exponentially as well. In the present times
Sentiment analysis is a field of study which
most of the people have smart phones and a
tries to identify the emotions or sentiments
high speed internet connection to go along
from the reviews of the customers.
with it.
Recent times has seen detailed research being
As a matter of fact, from Fig. 1 it can be seen
carried out in the field of sentiment analysis.
that in the year 2020 more than 4.79 Billion
Even though sentiment analysis is specifically
people are using internet, which is around
an application of Natural Language
61.4% of the world’s population . With India
Processing, the popularity and effectiveness of
alone accounting for more than 566 Million.
using Machine Learning and Deep Learning
This exponential growth in the internet users
techniques for conducting analysis of
has led to an unprecedented growth of data
sentiments has mandated a deeper study and
generated on the internet. This data is useless
research into the use of these learning
if relevant information is not extracted and
techniques for sentiment analysis.
used in decision making process. And,
Deep learning is a specialisation of machine
extraction of relevant information is very
learning. Deep learning techniques have been
crucial for corporations who are in the retail
found to be very effective for texts, images and
business space. The reason for this is, it has
videos. So, this paper proposes to conduct a
been observed that people are exceedingly
detailed critical review of deep learning
sharing their experiences on the products that
methods used for performing sentiment variants, Long Short Term Memory and Gated
analysis. Recurrent Unit. Qi Wang, et. al. propose to
experiment with these different variants of
recurrent neural network. The paper proposes
to experiment with all the different variants of
recurrent neural network, including long short
term memory and gated recurrent unit.
The paper proposes to make use of the IMDB
movie review dataset for experimentation.
Also, the paper makes use of GloVe for
converting the text into word vectors. After
experimenting with the variants of recurrent
neural network the paper then proposes to
experiment with using convolution layer as a
feature extraction layer. The paper claims that a
combination of convolution layer and bi-
Fig. 1: Worldwide social media users [1] directional long short term memory produces
the highest accuracy of around 89%.
1. Review of Deep Learning Techniques Lee Jun Sheng, et. al. propose develop a Gated
For Sentiment Analysis Recurrent Unit based Recurrent Neural
One of the important advantage of deep Network for Chinese texts. In order to do this
learning techniques over conventional machine the paper propose to develop dataset which
learning techniques is that deep learning would contain reviews on electronics items in
techniques can extract deeper intuition the Chinese language. The paper then performs
information in much better manner. This is manual labelling of this dataset.
probably the reason why deep learning Before applying the gated recurrent unit the
techniques have proved to be popular for paper performs a few natural language
performing research in sentiment analysis. processing tasks like tokenization and removal
This is because, proper feature extraction plays of stop words. The word embedding technique
a key role in sentiment classification accuracy. used has not been mentioned in the paper. The
And, as deep learning techniques perform paper claims to produce an accuracy of 87.6%.
feature extraction on their own and can On the other hand, Chayapol
extraction features with deep intuition, these Piyaphakdeesakun, et. al. aims to perform
techniques have generally produced very good sentiment analysis for opinions mentioned in
sentiment classification results. Thai language. In order to perform sentiment
While working with deep learning techniques analysis the paper proposes to experiment with
the architecture of the technique used, the type different combinations of deep learning
of dataset used and the type of word techniques. After experimentations the paper
embedding technique employed plays an concludes that for the developed dataset the
important role in determining the efficiency of accuracy result produced by Bi- Gated
the proposed system. So, among other things, Recurrent Unit based Recurrent Neural
the critical review has been carried out keeping Network are better at 91.85%, compared to
this in mind. other techniques.
All the different papers on sentiment analysis The paper proposes to develop a dataset on its
for deep learning can be categorized into five own by taking reviews from social media
types as under. platform. Which social media platform has
been primarily used for this data extraction has
1.1 Recurrent Neural Network and its not been mentioned in the paper. Also, the
Variants paper claims that different deep learning
One of the most important deep learning techniques have been experiment with but, the
technique used for sentiment classification has details about the implementation has not been
been the Recurrent Neural Network and its
provided as well. The type of word embedding embedding technique. The paper claims to
technique has not been mentioned either. have produced an accuracy of 87.86% for the
S.Anbukkarasi and S. Varadhaganapathy said deep learning model.
propose to analyse sentiments texts expressed Ru Ni, and Huan Cao are of the opinion that
in Tamil language. For this purpose the paper both the Long Short Term Memory and Gated
makes use of Bi-directional Long Short Term Recurrent Unit have different advantages
Memory based recurrent neural network. The which can be used appropriately. So, the paper
paper performs experimentation on simple propose to perform sentiment classification on
Long Short Term Memory and Bi-directional IMDB dataset using these two techniques in
Long Short Term Memory, it does not series that is, forst the Long Short Term
mentioned if any other deep learning model has Memory units have been employed and then its
been experimented with. output is fed to the Gated Recurrent Unit.
The paper mentions the importance of word The paper claims to have achieved an accuracy
embedding but, does not mention the word score of 87.1% by using the proposed model
embedding technique used in the proposed
with GolVe word embedding technique. It also
system. The Tamil scripted reviews for
sentiment classification has been collected and rather interesting to note that the same paper
manually labelled for the experimentation mentions that simple Long Short Term
purpose. The paper claims to have achieved an Memory produced an accuracy of 87.16%
accuracy of 86.2%. which is higher than the proposed model. If
On a similar lines even Rahma M. Alahmary this is true then it defeats the whole purpose of
propose to experiment with simple Long Short the system proposed in the paper.
Term Memory and Bi-directional Long Short Apart from the implementation of vanilla
Term Memory for sentiment classification. The recurrent neural networks, long short term
difference is that the paper tests the memory and gated recurrent are implimented.
effectiveness of Support Vector Machine as The reason for choosing these variants is the
well. Also, the paper performs sentiment training speed. Some of the papers have
analysis for Saudi dialect. implemented the bi directional versions of the
The paper proposes to collect reviews from same, one of the main reason for going for this
Twitter and create a dataset which would is the ability of bi directions versions to retain
contain texts in Saudi dialect. The paper gives contextual information over a large context.
a lot of attention at the creating of the dataset,
while also performing normalisation and 2.2 Convolution Neural Network
removing stop words. The paper makes use of Nguyen KhaiThinh, et. al. are of the opinion
continuous bag of words for word embedding that convolution layer with the gated recurrent
and claims to have an accuracy of 94%. unit produces better results. The paper
Whereas, Piyumal Demotte, et. al. are of the proposes to experiment with a combination of
view that using state information will be different deep learning techniques for
beneficial for proper sentiment classification. sentiment analysis on the IMDB movie review
So, the paper proposes to use State- Long Short dataset.
Term Memory for binary classification of The paper claims that after detailed
Sinhala text. The paper performs experiments experiments it observed that using of
on other deep learning techniques as well but, convolution layer before the gate recurrent unit
in the end shows that state- Long Short Term produces sentiment accuracy of 90.02%. The
Memory performs better for Sinhala text. paper does not mention the type of word
The paper proposes to use the word frequency embedding technique used of the said
experimentations. It can be assumed that the
based word embedding technique. Word
different use of word embedding is the only
frequency is one of the most basic form of reason for higher classification accuracy.
word embedding technique, the paper should
have gone with more state of the art word On the other hand, Fang Wan proposes to
experiment with effectiveness of using Deep
Neural Network in sentiment classification. paper claims to have experimented with

Even though the paper pledges to experiment different variants of recurrent neural network
with deep neural network, instead the paper and for a combination of Convolution Neural
makes use of Convolution Neural Network for Network and Bi-directional long short term
performing sentiment classification. memory in parallel the highest accuracy of
The paper focuses in the development of self-
82.14% was achieved.
trained word2Vec for word embedding. For
this paper proposes to make use of around Similarly, even Siddhartha Mukherjee propose
40000 reviews. The paper claims to have to work with a hybrid of Long Short Term
experimented with convolution neural network Memory based Recurrent Neural Network and
and produced a sentiment classification Convolution Neural Network in parallel. The
accuracy of 89.3%. But, the paper does not main aim of the paper though is to work with
mention the dataset used for the data which contains Hindi and English codes in
experimentation. a same dataset. So, the paper proposes to
The reason for specifically highlighting this is, develop a dataset which would contain Hindi
in machine learning and deep learning the type and English code for the training purpose.
and form of dataset plays a very important role
The paper proposes to make use of long short
in determine the accuracy of the system. This is
because, a deep learning model’s architecture term memory and convolution neural network
will have to be tweaked in order to produce in parallel. After this, the extracted features are
better results. concatenated for sentiment classification. The
There are very few papers which have paper mentions that word embedding has been
experimented with convolution neural network used by taking Devnagari script and the
for sentiment analysis. One reason for not English text but, does not mentioned which
going for convolution neural network is that it technique has been employed to train the word
cannot hold on to contextual information as embedding model. Also, the paper fails to
much as recurrent neural network can. But, on mention the dataset used for experimentations.
the other hand it is faster than recurrent neural
Finally, the paper claims an accuracy of
network, this is the main reason because of
which in the next category recurrent neural 69.85%.
network and convolution neural network. Similarly, even Li Yang, et. al. propose
perform sentiment analysis of reviews written
2.3 Recurrent Neural Network and
in the Chinese language. In order to perform
Convolution Neural Network
sentiment analysis the paper proposes to apply
Mehmet Umut Salur and Ilhan Aydin propose Convolution Neural Network and Bi-
to develop a hybrid system which makes the directional Gated Recurrent Unit based
best of both the worlds, that is, the paper Recurrent Neural Network in series. First, the
proposes to make use of both a full-fledged
Convolution Neural Network has been applied.
convolution neural network and a recurrent
neural network in parallel. The paper proposes The result of this layer has been given to the
to perform feature extraction at the character Bi-directional Gated Recurrent Unit. After both
level using the convolution neural network. these deep learning techniques finish operation
And, feature extraction at the word level using classification task has been performed.
the bi-directional long short term memory The paper mentions that the e-commerce
based recurrent neural network. reviews has been developed and used but, the
The paper proposes to perform these two tasks domain and the source of the dataset has not
in parallel. Then, both the features extracted been mentioned in the paper. The paper claims
are concatenated and classification is to have achieved an accuracy of around 93%.
performed. For the experimentation the paper It can be observed that when compared to only
makes use of the twitter reviews about mobile convolution neural network, when it is
network operators in Turkish language. The combined with recurrent neural network the
results obtained are comparatively higher. The combination of machine learning

technique and deep learning Technique has
2.4 Deep Learning and Machine
also produced good results. Most of the times
Learning
the Deep learning techniques are used as a
Punit Dubey, et. al. propose make use of Deep feature extraction step in machine learning
Neural predictions.
Network and Support Vector Machine for 2.5 Hybrid Techniques
sentiment analysis. The paper proposes to Siwei Lai, et. al. are of the opinion that while
make use of deep neural network as the feature performing sentiment analysis the
extraction step and the support vector machine understanding of the contextual information by
technique has been used as a classification step the learning is very important for more
for sentiment classification.
accurate classification. And, it is known that
The paper makes use of continuous bag of
recurrent neural network can hold on to the
words for embedding of the words. The paper
claims to have experimented with the said contextual information over a longer period
architecture but, it does not mention the dataset compared to convolution neural network,
used for the experimentation. Also, the paper which doesn’t deal with contextual
claim to have good results but, does not information. But, it is also true that the
mention the accuracy obtained. Because of this, performance of convolution neural network is
there is no way to ascertain the effectiveness of better than recurrent neural network.
the proposed architecture. So, the paper proposes to make use of both
On the other hand, Yogesh Chandra and these techniques. The paper proposes to
Antoreep Jana propose to make use of machine remove convolution layer in the convolution
learning technique, a polarity based technique neural network and replace it with bi-
and also, deep learning technique, all in directional recurrent unit. By doing this the
parallel. The paper is of the opinion that paper claims to not only hold on to the
different dataset will have different contextual information but also maintain the
characteristics and in some cases these efficiency of a convolution neural network.
characteristics are better handled by one of the The paper performs experimentations on
mentioned technique and not by other, and in different datasets to check the efficiency of the
some other cases it will be other way round. proposed model and claims to have produced
So, in order to cover all the scenarios the paper good accuracy scores. The highest accuracy
proposes make use of all these three score achieved is 96.46% for 20News dataset.
techniques. This is one of the highest results received even
The paper claims to have achieved an accuracy when used on multiple testing data.
of 97%. But, even though the paper mentions
the use of machine learning and polarity based 2. Analysis of Deep Learning
technique it does not give the details about the Techniques in Sentiment Analysis
actual technique employed in the proposed The papers reviewed can be generally
system. Only in with respect to the deep categorised into five distinct categories, as
learning aspect the paper mentions that Long shown in the Fig. 2. The first and most popular
Short Term Memory has been used in the category is the Recurrent Neural Network and
proposed system. its variants. In this categories all the different
Machine learning have also been independently variants of recurrent neural network has been
used, like or the. The results produced, for their included including the Long Short Term
respective dataset, ranges from 70s to 80s. But, memory and Gated Recurrent Unit, with their
it can be said that the introduction of deep respective bidirectional forms. The other
learning techniques increase the accuracy categories include convolution neural network,
scores. recurrent neural network and convolution
neural network applied in series, a combination • Sentiment Analysis of

of deep learning and machine chine learning Reviews Based on Deep
Learning Model [3]
techniques and a hybrid model. • Sentiment Analysis of
Chinese Product Reviews
It can be observed that the recurrent neural using Gated Recurrent Unit
[8]
network and its variants seem to be quite • Thai Comments Sentiment
Analysis on Social Networks
popular, and have been extensively used and with Deep Learning
researched for the purpose of sentiment RNN and Approach [10]
• Analyzing Sentiment in
analysis. One of the important reason for
f this Variants Tamil Tweets using Deep
Neural Network [12]
can be attributed to the retention of context in • Sentiment Analysis of Saudi
Dialect Using Deep Learning
recurrent neural network. For longer reviews Techniques [13]
understanding of the context plays an • Sentiment Analysis of
Sinhala News Comments
important in determining the actual user using Sentence
Sentence-State LSTM
Networks [14]
sentiments. • Sentiment Analysis based on
GloVe and LSTM
LSTM-GRU [15]
On the other side, while understanding and
retention of the contextual
ntextual information is very
• Sentiment Analysis Using
important, it is also true that the recurrent Residual Learning with
Simplified CNN Extractor [4]
neural networks and its variants are take a lot CNN • Sentiment Analysis of Weibo
of time for processing as the recurrent process Comments Based on Deep
Neural Network [5]
is a linear process. In order to mitigate this
drawback and speed up the processing • A Novel Hybrid Deep
Learning Model for
convolution
onvolution neural network have also been Sentiment Classification [6]
• Deep Learning Technique for
experimented with. But, convolution cannot Sentiment Analysis of Hindi
Hindi-
retain contextual information. RNN + English Code
Code-Mixed Text
using Late Fusion of
A solution to the problem of slow speed of CNN Character and Word Features
[7]
recurrent neural networks and non-retention
non of • Sentiment Analysis for E
E-
Commerce Product Reviews
contextual information has been found out by in Chinese Based on
Sentiment Lexicon and Deep
paper which proposes a hybrid model. This Learning [9]
hybrid model of applying the bi-directional
bi
• Sentiment Analysis Using
recurrent structure layer in place convolution SVM and Deep Neural
Network [11]
layer speed up the processing while retaining DL + ML • Sentiment Analysis using
the contextual information as well. Machine Learning and Deep
Learning [16]
Experimentations have been carried out by
• Recurrent Convolutional
applying
ying different algorithms together, like Hybrid Neural Networks for Text
done by the papers applying recurrent neural Classification [17]
network and convolution neural network in Fig. 2: Categories of Deep L Learning

series or by the papers who have used both the Techniques for Sentiment Analysis
machine learning and the deep learning The a deeper analysis of the above mentioned
algorithms. papers has been carried out in Tables 1, 2 and 3
on the basis of the type of deep learning
technique used, the dataset used in the paper
for experimentation, word embedd
embedding
technique used, the number classes included
for sentiment classification and finally the
accuracy of the system claimed in the paper.
Table 1: Analysis of deep learning techniques in sentiment analysis (A)

Sentiment Sentiment Analysis Sentiment A Novel Hybrid Deep Learning
Analysis of Using Residual Analysis of Weibo Deep Learning Technique for
Reviews Based Learning with Comments based Model for Sentiment Analysis of
on Deep Simplified CNN on Deep Neural Sentiment Hindi-English Code-
Learning Model Extractor [4] Network [5] Classification Mixed Text using Late
[3] [6] Fusion of Character
and Word Features [7]
Technique Convolution Convolution Neural Convolution Convolution Long Short Term
Used layer and Bi- Network Neural Network Neural Network Memory
directional Long and Bi-
Short Term directional Long
Memory Short Term
Memory in
parallel
Dataset used IMBD IMDB Not Mentioned Self-created Not mentioned
Turkish mobile
network twitter
dataset
Word GloVe Not mentioned Word2Vec Word2Vec Not mentioned
embedding
technique used
Number of 2 2 Not mentioned 3 Not mentioned
Classes
Accuracy 89% 90.02 89.3% 82.14% 69.85%
title must be capitalized except articles,

The title must be in 14 pt bold and be centered
prepositions, and conjunctions.
over both columns; the title should have 0 pts
space above and 12 pts below. ery word in a
Table.2. Analysis of deep learning techniques in sentiment analysis (B)

Sentiment Analysis of Sentiment Analysis Thai Comments Sentiment Analyzing
Chinese Product Reviews for E-Commerce Sentiment Analysis Using Sentiment in Tamil
using Gated Recurrent Product Reviews in Analysis on SVM and Deep Tweets using Deep
Unit [8] Chinese Based on Social Networks Neural Network Neural Network
Sentiment Lexicon with Deep [11] [12]
and Deep Learning Learning
[9] Approach [10]
Technique Gated Recurrent Unit Convolution Neural Bi-directional Deep Neural Bi-directional
Used Network and Bi- Gated Recurrent Network and Long Short Term
directional Gated Unit Support Vector Memory
Recurrent Unit Machine
Dataset used Self-created electronics Self-created e- Self-created Not mentioned Self-created
dataset in Chinese commerce dataset in dataset from
Chinese Twitter
Word Not mentioned BERT Not mentioned Continuous Bag of Not mentioned
embedding Words
technique used
Number of 2 Not mentioned 2 Not mentioned Not mentioned
Classes
Accuracy 87.6% 93% 91.85% Not mentioned 86.2%
Table.3. Analysis of deep learning techniques in sentiment analysis (C)

Sentiment Analysis of Sentiment Analysis Sentiment Sentiment Recurrent
Saudi Dialect Using of Sinhala News Analysis based Analysis using Convolutional
Deep Learning Comments using on GloVe and Machine Learning Neural Networks
Techniques [13] Sentence-State LSTM-GRU and Deep Learning for Text
LSTM Networks [15] [16] Classification [17]
[14]
Technique Used Bi-directional Long State- Long Short Long Short Machine Learning, Recurrent
Short Term Memory Term Memory Term Memory Polarity based Convolution
and Gated method and Long Neural Network
Recurrent Unit Short Term
in series Memory
Dataset used Self-created dataset News comments in IMDB First GOP Debate, 20News, Fundan,
from Twitter in Saudi Sinhala Bitcoin Tweets, ACL and SST
dialect and Live Twitter
Data
Word Continuous Bag of Word frequency GloVe Not mentioned Word2Vec
embedding Words
technique used
Number of Not Mentioned 2 2 2

Classes
Accuracy 94% 87.86% 87.1% 97% 96.49%
From the review it can be understood that 3. Conclusion

almost all the paper have performed multiple Deep learning techniques like the recurrent
experimentations on different deep learning neural network, convolution neural network
methods, this shows experimentations with and their variants have played a very
dataset is important. But on the other hand, promising role in the field of sentiment
hardly any paper gives details about the analysis. These techniques improved the
architecture of the deep learning model efficiency of the systems while performing
developed. They do not mention the details analysis of sentiments.
about the number of layers nor do they Understanding the importance of these deep
divulge any details about the hyper learning techniques in sentiment analysis, a
parameters values used. critical review has been carried out in this
Also, either only two classification classes paper. From the detailed study it can be
have been used for experimentation or the concluded that in sentiment analysis the
paper do not mention the number of classes understanding and storing of contextual
taken at all. As the number of classes information plays a very crucial role in
increases more there is more strain on the determining the efficiency of the system in
deep learning model and the accuracy score is classifying the sentiments.
adversely affected. Even though word Even then, it can be seen that quite a few
embedding is an essential and an important papers so not fully appreciate the importance
task before applying any deep learning of the model having the contextual
technique, it has been observed that not much information of the text under consideration
importance has been given to word for sentiment classification. Basic recurrent
embedding technique. This is because, most neural network, long short term memory and
of the paper do not mention the technique gated recurrent unit do hold contextual
used of word embedding at all. It should be information but, this information is lost over a
noted here that the same model will produce large text. It is also true that these variants of
different accuracy scored for different dataset. recurrent neural network takes a lot of time
A model performing on one dataset might fail for training and testing that is, they are not
miserably on another dataset. efficient. So, it can be said that the paper
which proposes to not only hold contextual across different dataset is an important
information but also, produce results in an contribution in the field of sentiment analysis.
efficient manner with very good accuracy
References
1. C. Piyaphakdeesakun, N. Facundes, and J. Classification”, IEEE Access, Vol. 8, 2020,

Polvichai, “Thai Comments Sentiment pp. 58080-58093.
Analysis on Social Networks with Deep 9. N. KhaiThinh, C. H. Nga, Y. S. Lee, M. L.
Learning Approach”, IEEE 34th Wu, P. C. Chang, and J. C. Wang,
International Technical Conference on “Sentiment Analysis Using Residual
Circuits/Systems, Computers and Learning with Simplified CNN Extractor”,
Communications (ITC-CSCC), 2019. IEEE International Symposium on
2. Editors, “Key Internet Statistics to know in Multimedia (ISM), 2019, pp.335-338.
2020 (Including Mobile)” [Online]. 10. P. Demotte, L. Senevirathne, B.
https://www.broadbandsearch.net/blog/inte Karunanayake, U. Munasinghe, and S.
rnet-statistics, Last accessed in 20th July, Ranathunga, “Sentiment Analysis of
2020. Sinhala News Comments using Sentence-
3. F. Wan, “Sentiment Analysis of Weibo State LSTM Networks”, IEEE Moratuwa
Comments Based on Deep Neural Engineering Research Conference
Network”, IEEE International Conference (MERCon), 2020, pp. 283-288.
on Communications, Information System 11. P. Dubey, A. Mishra and B. K. Saha,
and Computer Engineering (CISCE), 2019, “Sentiment Analysis Using SVM and Deep
pp. 626-630. Neural Network”, IEEE Third International
4. L. J. Sheng, D. Zuba and P. Yan, Conference on Computing Methodologies
“Sentiment Analysis of Chinese Product and Communication (ICCMC 2019), 2019,
Reviews using Gated Recurrent Unit”, pp. 952-957.
IEEE Fifth International Conference on Big 12. PTI. (2019, Mar 06). Internet users in India
Data Computing Service and Applications to reach 627 million in 2019: Report
(BigDataService), 2019, pp. 173-181. [Online].
5. L. Yang, Y. Li, J. Wang, and R. S. Sherratt, https://economictimes.indiatimes.com/tech/
“Sentiment Analysis for E-Commerce internet/internet-users-in-india-to-reach-
Product Reviews in Chinese Based on 627-million-in-2019-
Sentiment Lexicon and Deep Learning”, report/articleshow/68288868.cms?from=m
IEEE Access, Vol. 8, 2020, pp. 23522- dr, Last accessed 16th Oct, 2019.
23530. 13. Q. Wang, L. Sun and Z. Chen, “Sentiment
6. M. H. Abd El-Jawad, R. Hodhod and Y. M. Analysis of Reviews Based on Deep
K. Omar, “Sentiment Analysis of Social Learning Model”,IEEE/ACIS 18th
Media Networks Using Machine International Conference on Computer and
Learning”, IEEE 14th International Information Science (ICIS), 2019, pp. 258-
Computer Engineering Conference 261.
(ICENCO), 2018, pp. 174-176. 14. R. M. Alahmary, H. Z. Al-Dossari, and A.
7. M. Rumelli, D. Akkuş, Ö. Kart and Z. Isik, Z. Emam, “Sentiment Analysis of Saudi
“Sentiment Analysis in Turkish Text with Dialect Using Deep Learning Techniques”,
Machine Learning Algorithms”, IEEE IEEE International Conference on
Innovations in Intelligent Systems and Electronics, Information, and
Applications Conference (ASYU), 2019, Communication (ICEIC), 2019.
pp. 1-5. 15. R. Ni, and H. Cao, “Sentiment Analysis
8. M. U. Salur and I. Aydin, “A Novel Hybrid based on GloVe and LSTM-GRU”, IEEE
Deep Learning Model for Sentiment 39th Chinese Control Conference, 2020,
pp. 7492-7497.
16. S. Anbukkarasi and S. Varadhaganapathy, AAAI Conference on Artificial

“Analyzing Sentiment in Tamil Tweets Intelligence, 2015, pp. 2267-2273.
using Deep Neural Network”, IEEE Fourth 19. S. Mukherjee, “Deep Learning Technique
International Conference on Computing for Sentiment Analysis of Hindi-English
Methodologies and Communication Code-Mixed Text using Late Fusion of
(ICCMC 2020), 2020, pp. 449-453. Character and Word Features”, IEEE 16th
17. S. Hemalatha and R. Ramathmika, India Council International Conference
“Sentiment Analysis of Yelp Reviews by (INDICON), 2019.
Machine Learning”, International 20. Y. Chandra and A. Jana, “Sentiment
Conference on Intelligent Computing and Analysis using Machine Learning and Deep
Control Systems (ICCS), 2019, pp. 700-704 Learning”, IEEE 7th International
18. S. Lai, L. Xu, K. Liu, and J. Zhao, Conference on Computing for Sustainable
“Recurrent Convolutional Neural Networks Global Development (INDIACom), 2020,
for Text Classification”, Twenty-Ninth pp. 1-4.
OPTIMIZED 2DOF-PIDN CONTROLLER BASED AGC SYSTEM OF A MULTI SOURCE

THERMAL-HYDRO-WIND INTERCONNECTED RESTRUCTURED POWER SYSTEM
E. Prakash and I. A. Chidambaram
Department of Electrical Engineering, Annamalai University, Annamalainagar, Tamilnadu, India
elamprakash@gmail.com
_______________________________________________________________________________________
ABSTRACT
This paper proposes the two degree of freedom Proportional-Integral-Derivative with derivative Filter (2DOF-PIDN)
controller based Automatic Generation Control (AGC) of a two-area multi-source thermal-hydro-wind interconnected
restructured power system. A two degree of freedom controllers provides an output which is difference between a
reference signal and a measured output of system and performs better in set point tracking and provides better
regulation for disturbance input as compared with single degree of freedom controllers. The fundamental points of
interest of the Proportional-Integral-Derivative (PID) controller are that a derivative mode improves the stability of
the system. However, the derivative term will create the awkward size of control inputs to the plant. Also, it
produces/propagates noise along with the control input signal resulting in large plant input signals. The practical
solution for these issues is to put the first-order filter on the derivative term and tune its pole so that the controller
response does not produce noise since it attenuates high-frequency noise. In this examination, a 2DOF-PIDN
controller are designed with filter in the derivative term and implemented as supplementary controller of AGC loop of
a two-area thermal-hydro-wind interconnected restructured power system. The control parameters of the 2DOF-PIDN
controller are tuned using Buzzard Optimization Algorithm (BUZOA) and its performance is compared with PI and
PIDN controller. Further, this study has been carried out to evaluate power system Ancillary Service Requirements
(ASR) indices based on the AGC assessment are processed relies upon the dynamic response of the control input
deviations and change in power generation of Gencos of each area. Simulation results show that the proposed BUZOA
tuned 2DOF-PIDN controller improves the dynamic output response of the test system in terms of less peak deviation
and settling time of area frequencies and tie-line power oscillations of the a two-area multi-source thermal-hydro-wind
interconnected restructured power system as compared with PI and PIDN controller and guarantee to improved ASR
indices so as to upgrade the system stability.
________________________________________________________________________________
Keywords: AGC, Buzzard Optimization Algorithm, PI, PIDN, 2DOF-PIDN controller
1. Introduction problem of AGC can be classified into fast

primary control and slow secondary control
The essential responsibility of Automatic modes. The fast primary control (governing
Generation Control (AGC) is used to maintain mechanism) mode tries to reduce the frequency
the system frequency associated with tie-line deviations along with the time constant in the
power deviations. These deviations are the order of seconds. But, primary control does not
roots of deviations with generation and load assure the zero steady-state error. The slow
demand. On this occasion the deviations are secondary control channel (supplementary
kept unnoticed, at that point, it might prompt control), with time constants in the order of
severe undesirable impacts. In the current minutes, normalizes the generation to gratify
situation, the power system course of action some loading requirements and contractual tie-
comprises the mass of generating utilities line loading agreements. The best performance
interconnected together through tie-lines. The of the AGC in any power system stands on the
advanced power system usually comprises of proper design of both primary and secondary
various interconnected subsystems with tie- control loops.
lines. For each part of the subsystem is The AGC action is synchronized by the Area
necessitates to incorporating the matching Control Error (ACE) with the aid of system
system generation with system load used to frequency and tie-line flows. In every zone,
control the system frequency. Restoration of AGC monitors the system frequency and tie-
system frequency used with ostensible value line flow and works out the net change in the
needs supplementary control action which generation required based on the change in
regulates the load reference set point through demand along with the changes in the set
the speed changer mechanism. Therefore, the position of the generators within the area used
to keep the time average of the ACE at a low unity. DPM proves the participation of a Disco
value. In this manner, ACE is described as a in a contract with Genco.
linear combination of power net-exchange and The nominal operating point of a power system
frequency deviations considered as the transforms from its
controlled output of AGC. As the ACE is predetermined value subjected to any
headed to zero by the AGC, frequency and tie- disturbance due to the deviation made by the
line power errors will be compelled to zeros. operating point in the ostensible system
The regular AGC two-area interconnected frequency and scheduled power trade to
power system is altered to consider the activity different areas which is unfortunate. Thus, an
of AGC in an open market power system. Open AGC scheme primarily integrates a suitable
transmission with other socialized companies control system for an interconnected power
for generation, transmission, and distribution system having the ability to bring the
affects the design of the AGC problem to frequencies of every area and the tie-line power
house the new constraints along with the back to unique set point values or very closer
territorial functionality of each company. So to set point values successfully after the load
the usual AGC two-zone interconnected power change. Robust secondary controllers are
system is manipulated in accordance with the important to retain a flat frequency profile.
effect of bilateral contracts on the dynamics. Several advanced controller structures and
In perspective on the bilateral transactions, a techniques have been proposed in the literature
distribution company (Disco) has the chance to for AGC. Among the choice of secondary
contract with any open generation company controllers, the most commonly employed is
(Genco) in its own one of a kind or another the conventional Proportional plus Integral (PI)
control zone. In this manner, the idea of the controllers. These PI controllers are still
physical control region is supplanted by Virtual popular in the power industry for frequency
Control Area (VCA). The limit of a VCA is regulation even in any change in system
adaptable and encases the Gencos and the operating conditions, new gain values can be
Disco related to the agreement. These computed easily even for multi-area power
examinations endeavor to correct the usual system. The PI controller is very simple for
AGC system used to reflect on the impact of execution and gives a better dynamic response,
bilateral contracts on the dynamics, improve but their performances deteriorate when the
the dynamic and transient response of the complexity in the system increases due to the
system under different working conditions. natural impact of the disturbances. The
Each control territory has its very own AGC fundamental downside of this PI controller is
and is in charge of following its own load and doesn't capable to predict the future errors of
respecting tie-line power trade contracts with the system, can't eliminate steady-state
its neighbors. Starting at now, these trades are oscillations and reduces settling time.
done under the supervision of the Independent Consequently, the overall stability system is
System Operator (ISO). There can be relatively low.
distinctive combines the understandings The fundamental points of the Proportional
between each Disco and available Gencos, on plus Integral plus Derivative (PID) controller, a
the other hand, each Genco can contract with derivative mode improves the stability of the
various Discos. With the development of the system and facilitates to increase in
Disco Participation Matrix (DPM) the proportional gain and decreases in an integral
visualization of contracts can be made easier. gain which thusly expands the speed of the
DPM is represented as an asymmetrical matrix controller response. When the input signal has
constructed by the number of rows equal to the sharp corners, the derivative term will deliver
number of Gencos and the number of columns unreasonable size control inputs to the plant.
equal to the number of Discos in the system. Additionally, any noise in the input signal will
Each element in the matrix is referred to a bring about enormous plant input signals that
fraction of a total load contracted by a Disco lead to confusion in pragmatic applications.
(Column) towards a Genco (Row). The sum of The practical solution for this issue is forcing
all the entries in a column of this matrix is the first-order filter on the derivative term and
tunes its pole so that the chattering due to the 2.1 Control structure of 2DOF-
noise does not occur since it attenuates high- PIDNcontroller
frequency noise. A Proportional Integral The degree of freedom of a control system is
Derivative with derivative Filter (PIDF) defined as the number of closed-loop transfer
controller had been considered in that study. functions that can be adjusted independently.
From literature survey the enhancement of While designing a control system various
power system performance not only depends performance criteria are to be satisfied so a
on the control structure but also on the well- two-degree-of-freedom control system
tuned controllers. This paper proposes the two naturally has advantages over the conventional
degree of freedom Proportional-Integral- single degree of freedom control system. The
Derivative with derivative Filter (2DOF-PIDN) 2-DOF controller produces an output signal
controller based Automatic Generation Control based on the difference between a reference
(AGC) of a two-area multi-source thermal- signal and a measured system output. It
hydro-wind interconnected restructured power computes a weighted difference signal for each
system. A two degree of freedom controllers of the proportional, integral and derivative
provides an output which is difference between actions according to the specified set point
a reference signal and a measured output of weights. The controller output is the sum of the
system and performs better in set point tracking proportional, integral and derivative actions on
and provides better regulation for disturbance the respective difference signals where each
input as compared with single degree of action is weighted according to the chosen gain
freedom controllers. The control parameters of parameters. The structure of proposed parallel
the 2DOF-PIDN controller are tuned using 2-DOF PIDN controller is shown in Fig.1,
Buzzard Optimization Algorithm (BUZOA) where R(s) represents the reference signal, Y(s)
and its performance is compared with PI and represents the feedback from measured system
PIDN controller. Buzzard Optimization output and U(s) represents the output signal,
Algorithm (BUZOA) is one of the modern KP, KI and KD are the proportional, integral and
heuristic algorithms and has a great potential to derivative gains respectively, PW and DW are
solve complex optimization problems. the proportional and derivative set point
2. Design of 2DOF-PIDN Controllers Using weights respectively, and N is the derivative
Buzzard Optimization Algorithm filter coefficient. A block diagram of 2-DOF
control system is shown in Fig.2.
Fig. 1 Block diagram for 2 DOF-PIDN controller
2.2 Buzzard Optimization Algorithm based

design of 2DOF-PIDN controller
Buzzard Optimization Algorithm (BUZOA) is
one of the modern heuristic algorithms and has
a great potential to solve complex optimization
Fig. 2 Block diagram for 2 DOF control system problems. This algorithm is modeled by
paraphrasing and simulated with the behavior
In this Fig 2, C(s) is a single degree-of-freedom
of collective flight (group) buzzards (vultures).
controller, D(s) is the load disturbance and F(s)
A piece member in same group is presented by
acts as a pre filter on the reference signal. For a
the ability vector and the position vector in all
parallel two-degree-of-freedom PIDN
the search space. The time iteration has
controller, C(s) and F(s) are given by:
presented the new position of a particle by the
( ) ( ) position vectors and the ability vectors in the
( )= ( ) ( )
(1)
search space. The new and next position of a
( ) ( ) particle, regarding the current ability vector,
( )= (2)
( ) and the best position founded by that particle
The main function of the AGC system is to and the best particle in the same group in the
control the load-frequency and tie-line power turkey buzzards position is updated in the each
during load disturbance. So the error signal of iteration. Let us assume d -dimensional
frequency and tie-line power are utilized as buzzard search space. i -th particle is
design criteria to tune the 2DOF-PIDN described by the vector of position Li as follows
controller. This error signal is used to produce in this d -dimensional space:
the individual ACE signal as in the regular = , ,…… (8)
,
circumstance.
C -vector is the ability of smell, and the ability
= ∆ +∆ (3) of taste for the i -th particle is defined by the
= ∆ + ∆ (4) vector Ci as follows:
The error inputs to the controllers are the = ,, ,, ,, … … (9)
individual ACE signals is given by Eqn. (3) The best position which i -th particle is found
and (4). The BUZOA technique is used to is thevector ∗, and are shown as follows:
decide the optimal parameters of 2DOF-PIDN
∗ ∗ ∗ ∗ ∗
controllers with the goal to limit integral square , =( , , ,…… , ) (10)
of ACE signal, which can be detailed in an
The best position that has found the best
accompanying way: The performance of these
particle inthe whole particle is ∗, and
responses is measured using performance
functions such as Integral of Squared Error defines as follows:
(ISE) given by ∗ ∗ ∗ ∗ ∗
, = , , ,…… , (11)
=∫ [( ) ] (5) The best position in the whole particle is
∗
, , achieve the best position of each
The problem constraints are the proposed
controller parameter bounds. Therefore, the particle ∗, and best position among all
∗
design problem can be formulated as, particles , , we select the position of each
particle as the best location achieved for that
(6) particle in the first iteration ( =1).
Subject to ∗
= ( ), = 1, 2, 3, … (12)
,
≤ ≤ , ≤ ≤ ∗
( , )= ( ( )) (13)
, ≤ ≤ , ≤ ≤
, ≤ ≤ , ≤ ≤ The changing location and cost of each particle
(7) in the process iteration algorithm as follows.
( ) < ( ∗
⇒ location, we limit the variation of the ability to
⎧ ,
⎪ the range Cmin ≤ C ≤ Cmax.. is smell
ℎ capability, is the vision capability and smell
⎨ ∗
( , )= ( ) = 1, 2, 3 … = 1-taste. The flowchart of Buzzard
⎪ ∗ Optimization Algorithm is depicted in Fig.3.6.
⎩ , = ( )
The ACE minimization for optimal estimations
(14) of controller’s coefficients has been tackled
To update the location of each particle, the utilizing the BUZOA technique. The general
followingequations are used: straight forwardness of this controller is a
∗ fruitful methodology towards the optimal
( )= ( − 1) + ∗ ∗( , −
transient response of the system under different
( − 1)) (15) sorts of possible exchanges in the restructured
( )= ( )+ ∗ ∗ ( )− condition.
−1+1− ∗ 2∗ − −1+ ∗ 1∗ 3. Modeling of a 2DOF-PIDN Controller
∗
, − ( − 1) + (1 − ) ∗ ∗ Based Agc Loop of a Two-Area Multisource
∗
Interconnected Thermal-Hydro-Wind
, − ( − 1) (16) Restructured Power System
The power systems are integrating increasing
levels of variable renewable energy resources
( )= ( − 1) + ( ) (17) in the electric power mix. Currently, many
: the inertia weighting factor, used in the winds systems exist throughout the world and
insider path and , and are define the supply a significant portion of energy to the
training constant coefficient. , are networks. Because variable renewable energy
two random numbers with uniform distribution generation is variable and uncertain power
in interval 0 to 1. Updating the ability vector systems with increasing shares of variable
each particle is defined in term of bellow renewable energy require more flexibility to
equation. balance electricity supply and demand at all
∗
timescales. One measure of power grid
( )= ( − 1) + ∗ ∗( , − reliability is the stability of a system’s
( − 1)) (18) frequency at a desired level. This study
( )= ( )+ ∗ ∗ ( )− addresses the impact of hydro and wind power
generation on the AGC system of a two-area
−1+1− ∗ ∗ − −1+ ∗ ∗ ,
multi-source restructured power system. Each
∗− −1 (19)
area has two Gencos and Discos. The Gencos
in area-1 is a thermal reheat power plant unit
and hydro power plant unit and area-2 is a
( )= ( − 1) + ( ) (20) thermal reheat power plant unit and wind
( − 1), is the ability vectors in repetition power plant unit. Transfer function block
( − 1)th and ( − 1) is the Position vector in diagram model of 2DOF-PIDN controller
repetition( − 1)th. To prevent the excessive based AGC system for a two-area multi-source
increase in the ability and speed of a particle, in thermal-hydro-wind interconnected
the movement from one location to another restructured power system is shown in Fig.3.
Fig. 3 Transfer function block diagram model of 2DOF-PIDN controller based AGC system for a
two-area multi-source thermal-hydro-wind interconnected restructured power system
4. Evaluation of Power System Ancillary Gencos of each area for different sorts of
Service Requirements (ASR) Indices possible exchanges. The different power
system Ancillary System Restoration indices
Power system restoration is all around
(ASR1, ASR2, ASR3, ASR4, ASR5 and ASR6)
perceived as a significant undertaking to
are determined as pursues
diminish the effect of a disturbance that
Stage 1: The ASR1 is obtained from the
happens in power systems. The significant
deviation of dynamic response with the change
level methodology of the systems restoration
in control input of area-1 at peak time from the
plan is to reestablish the respectability of the
final value of the response
interconnection as fast as could be expected
under the circumstances. The system =∆ −∆ (∞) 21)
restoration procedures are found firmly
identified with the system characteristics. In Stage 2: The ASR2 is obtained from the
the wake of investigating the conditions and deviation of dynamic response with the change
characteristics of blackouts, system restoration in control input of area-2 at peak time from the
organizers or dispatchers will choose the final value of the response
appropriate remedial measures using the power =∆ −∆ (∞) (22)
system ASR indices which were obtained with
the dynamic response of the control input Stage 3: The ASR3 is acquired from the
deviations and change in power generation of deviation of dynamic response with the change
in power generation of Genco1 to reach
the peak value for the first time from the final Participation Matrix (DPM) alluding is
value of the response considered as
=∆ −∆ (∞) (23) 0.4 0.5 0.2 0.2
0.3 0.3 0.1 0.1
Stage 4: The ASR4 is acquired from the = (27)
0.2 0.1 0.5 0.4
deviation of dynamic response with the change 0.1 0.1 0.2 0.3
the peak value for the first time from the final For this situation, each Disco demands 0.25
value of the response pu.MW for each from Gencos as characterized
by cpf in the DPM matrix and each Gencos
=∆ −∆ (∞) (24) takes part in AGC as characterized by the
Stage 5: The ASR5 is acquired from the accompanying area participation factor apf11
deviation of dynamic response with the change = apf12 = 0.5 and apf21 = apf22 = 0.5. The
in power generation of Genco3 to reach BUZOA technique is used to optimize PI,
the peak value for the first time from the final PIDN and 2DOF-PIDN controller gain
value of the response estimations of the test system is shown in
Table 1. These controllers are executed in a
=∆ −∆ (∞) (25) proposed test system and compared with that of
the PI and PIDN controllers. The comparative
Stage 6: The ASR6 is acquired from the
dynamic output responses of the area
deviation of dynamic response with the change
frequencies and tie-line power deviations of the
test systems with various sorts of controllers
the peak value for the first time from the final
have shown up in Fig.4. The setting time and
value of the response
peak over/under shoot of the area frequencies
=∆ −∆ (∞) (26) and tie-line power deviations of the test
systems using PI, PIDN and 2DOF-PIDN
5. Simulation Results and Observations controllers have appeared in Table 2. From the
In this examination AGC performance of a Fig 4 and Tables 2, it very well may be seen
two-area multi-source thermal-hydro-wind that the proposed 2DOF-PIDN controller have
power generating units of restructured power better dynamic responses of frequency
system using PI/PIDN/ 2DOF-PIDN deviations of each area and tie-line power
controllers is considered. The model of the deviations as far as settling time and peak
power system under investigation has been over/under shoot are concerned when
developed in the matlab / simulink compared with that of PI/ PIDN controllers.
environment. The parameters values of the The above analysis revealed that 2DOF-PIDN
two-area multi-source AGC system are given controller has less peak deviation, magnitude
in the appendix. The proposed secondary of oscillations, and faster settling time than PI/
controllers like PI, PIDN and 2DOF-PIDN are PIDN controllers and shows superior
considered one at a time. The control performance for controlling of system
parameters and gain values of the proposed PI/ oscillations.
PIDN/ 2DOF-PIDN controllers are optimized The Comparison of dynamic responses of the
through BUZOA technique with the goal to change in control input deviations and change
limit the ACE signal for the two-area multi- in power generation of each Genco for the test
source thermal-hydro-wind restructured power. system using PI, PIDN and 2DOF-
The optimal solution of control inputs is taken PIDN controller is shown in Figs 5 and 6.
an enhancement issue and the target work in
From the Figs 5 and 6, the different power
Eqn (5) is inferred utilizing the frequency
deviations of control areas and tie-line power system Ancillary System Restoration indices
changes. In the bilateral exchanges every one (ASR1, ASR2, ASR3, ASR4, ASR5 and ASR6)
of the Discos has an agreement with the are determined using Eqn (21-26) of the test
Gencos and the accompanying Disco system with different PI / PIDN / 2DOF-PIDN
controllers. The ASR indices of PI/ PIDN/
2DOF-PIDN controller based AGC system for dynamic responses of control input deviations
the test system are recorded in the Table 3. and change in power generation of Genco in
From the Table 3, ASR indices the restorative each area of the test system under various
measures like the change in control input uncontract step load demand change
deviations and change in power generation of conditions. Furthermore, the proposed 2DOF-
Genco in each area can be adjudged. Likewise, PIDN controller-based two-area thermal-
the dynamic execution and ASR indices reflect hydro-wind ensures improved ASR indices to
that the 2DOF-PIDN controller based AGC provide reduce the restoration time, thereby
loop of the test system had improved the improving the system reliability.
Table 1 Optimal control parameters of the for the test system

Controllers Gain and control parameters Area-1 Area-2
PI KPi 0.439 0.277
KIi 0.243 0.251
PIDN KPi 0.311 0.194
KIi 0.403 0.342
KDi 0.589 0.401
N 8.341 6.031
2DOF-PIDN KPi 0.532 0.423
KIi 0.457 0.371
KDi 0.643 0.617
N 16.77 14.79
PWi 0.525 0.733
DWi 0.745 0.953
Fig. 4 (a) Dynamic responses of the frequency deviations of Area-1
Fig. 4 (b) Dynamic responses of the frequency deviations of Area-2
Fig. 4 (c) Dynamic responses of the Tie-line power deviations (actual)
Fig. 4 (d) Dynamic responses of the Tie-line power deviations (error)

Fig. 4 Comparison of dynamic responses of the power deviations for the test system using PI,
change in frequency deviations and tie-line PIDN and 2DOF-PIDN controller
Fig. 5 (a) Dynamic responses of the control input deviations of Area-1
Fig. 5 (b) Dynamic responses of the control input deviations of Area-2

Fig. 5 Comparison of dynamic responses of the system using PI, PIDN and 2DOF-PIDN
change in control input deviations for the test controller
Fig. 6 (a) Dynamic responses of change in power generation of Genco-1
Fig. 6 (b) Dynamic responses of change in power generation of Genco-2
Fig. 6 (c) Dynamic responses of change in power generation of Genco-3
Fig. 6 (d) Dynamic responses of change in power generation of Genco-4

Fig. 6 Comparison of dynamic responses of the test system using PI, PIDN and 2DOF-PIDN
change in power generation of Genco for the controller
Table 2 Dynamic performances of area frequencies and tie-line power deviations of test
system using PI/ PIDN/2DOF-PIDN controllers
Controllers Settling time of test system Peak over/under shoot of test system
F1 in sec F2 in sec Ptie in sec F1 in Hz F2 in Hz Ptie in p.u.MW
PI 36.371 34.523 35.084 2.052 1.503 0.403
PIDN 29.743 30.301 26.804 1.806 1.368 0.361
2DOF-PIDN 26.804 24.957 25.518 1.497 1.105 0.282
Table 3 ASR indices of AGC loop for the test system using PI/ PIDN/2DOF-PIDN controllers
Controllers ASR indices of AGC loop for the test system
ASR1 ASR2 ASR3 ASR4 ASR5 ASR6
PI 0.371 0.299 0.367 0.501 0.356 0.479
PIDN 0.307 0.265 0.351 0.477 0.341 0.458
2DOF-PIDN 0.271 0.206 0.303 0.438 0.312 0.416
6. Conclusion based test system ensures improved ASR

indices which give an extraordinary edge of
The proposed 2DOF-PIDN controllers are
robustness. These ASR indices indicate that the
designed utilizing BUZOA and implemented in
ancillary help necessary to improve the
a two-area multi-source thermal-hydro-wind
effectiveness of the physical activity of the
restructured power system. Investigation
power system with the expanded transmission
reveals that the comparison of dynamic
limit in the network. The proposed 2DOF-
responses the proposed BUZOA tuned 2DOF-
PIDN controller's based AGC system for two-
PIDN controllers exhibit quick, progressively
area thermal-hydro-wind interconnected
precise and superior dynamic characteristics
restructured power system shows better
than that of the responses obtained using PI/
execution to guarantee improved ASR indices
PIDN controllers. It is also observed that the
so as to upgrade the system stability.
frequency deviations of every area and tie-line
power deviations have been diminished 7. Acknowledgement
regarding fewer peaks over / under shoot and The authors wish to thank the authorities of
decreased settling time as compared with PI / Annamalai University, Annamalainagar,
PIDN controllers. Furthermore, the restoration Tamilnadu, India for the facilities provided to
procedure for the 2DOF-PIDN controller's prepare this paper.
References
1. Arshaghi, A., Ashourian, M. and Ghabeli, 6. Pappachen, A. and Fathima, A.P., (2017),
L., (2019), Buzzard optimization Critical research areas on load frequency
algorithm: a nature-inspired metaheuristic control issues in a deregulated power
algorithm, Majlesi Journal of Electrical system: A state-of-the-art-of-
Engineering, Vol.13 (3), pp.83-98. review, Renewable and Sustainable
2. Baskar, B., Paramasivam, B., and Energy Reviews, Vol.72, pp.163-177.
Chidambaram, I. A., (2018), Ancillary 7. Pardeep Nain, Singh Parmar, K.P. and
Service Requirement based Automatic Singh, A.K., (2013) “Automatic
Generation Control Assessment in a Generation Control of an Interconnected
deregulated power system with HES and Power System Before and After
IPFC units, Journal of Engineering Deregulation”, International Journal of
Science and Technology, Vol.13(10), Computer Applications, Vol. 61 (15),
pp.3092-3115 pp.11-16.
3. Chidambaram, I.A. and Paramasivam, B., 8. Ravi Shankar, Pradhan, S.R., Kalyan
(2012), Control performance standards Chatterjee and Rajasi Mandal,, “A
based load-frequency controller comprehensive state of the art literature
considering redox flow batteries survey on LFC mechanism for power
coordinate with interline power flow system”, Renewable and Sustainable
controller, Journal of power Energy Reviews, Vol. 76, pp.1185–1207,
sources, Vol.219, pp.292-304. 2017.
4. Donde, V., Pai, M.A. and Hiskens, I.A., 9. Sadia Firdous, Mohammed Abdul
(2001), Simulation and optimization in an Rahman Uzair, (2015), Load Frequency
AGC system after deregulation, IEEE Control Techniques, International Journal
transactions on power systems, Vol.16 of Science, Engineering and Technology
(3), pp.481-489. Research, Vol. 4(8), pp.2887-2893.
5. Fosha, C.E. and Elgerd, O.I., (1970), The 10. Sahu, R. K., Panda, S., and Padhan, S.,
Megawatt Frequency Control problem: A “Optimal gravitational search algorithm
New Approach via Optimal Control for automatic generation control of
Theory, IEEE Transactions on Power interconnected power systems”, Ain
Apparatus and Systems, Vol. PAS-89 (4), Shams Engineering Journal, Vol.5 (3),
pp. 563-574. 721-733, 2014.
11. Sahu, R. K., Panda, S., Biswal, A., and area interconnected thermal system”,
Sekhar, G. C., “Design and analysis of tilt International Journal of Electrical Power
integral derivative controller with filter & Energy Systems, Vol.33 (3), pp.394-
for load frequency control of multi-area 401, 2011.
interconnected power systems”, ISA 13. Sánchez, J., Visioli, A. and Dormido, S.,
transactions, Vol.61, pp.251-264, 2016. (2011), A two-degree-of-freedom PI
12. Saikia, L.C., Nanda, J. and Mishra, S., controller based on events, Journal of
“Performance comparison of several Process Control, Vol.21 (4), pp.639-651.
classical controllers in AGC for multi-
9. Appendix
Table.1: Control area, Gencos parameters [3, 10]

Parameters Area1 Area 2
Area capacities 1000 MW 1000 MW
Rating of single generating machine 500 MW 500 MW
Kp (Hz/p.u.MW) 120 120
Tp (sec) 20 20
β (p.u.MW / Hz) 0.425 0.425
R (Hz / p.u.MW) R1=R2=2.4 R3=R4=2.4
Tg (sec) Tg1= 0.08 Tg3 = 0.08
Tt (sec) Tt1= 0.36 Tt3 = 0.36
Tr(sec) Tr1= = 10 Tr3= 10
Kr Kr1= 0.5 Kr3= 0.5
Synchronising coefficient (p.u.MW / Hz) 2T12=0.545
System frequency (F) in Hz 60 Hz
Area participation factor (apf) apf11=apf12=apf21=apf22= 0.5
Area capacity ratios a12 = -1
Hydro power generating unit THg =0.2 sec, T1=0.513 sec,
T2=10 sec, Tw=1 sec,
Wind power generating unit Tw1=6.0 sec, Tw2=0.041 sec
Kw1= 1.25, Kw2=1.4
OPTIMIZED FREQUENCY AND POWER CONTROL STRATEGY FOR HYBRID

ENERGY SYSTEM
K. Kalpana and M.M.
M. T. Ansari
Department of Electrical Engineering, Annamalai University, Annamalainagar, India
______________________________________________________________________
_______________________________________________________________________________________
ABSTRACT
Power and frequency stabilization arethe major factorsthat should be considered in hybrid energy systems to improve
system efficiency. Due to economic reasons, wind-diesel
wind diesel hybrid energy systemsare widely adopted. The voltage and
frequency of the system depend on the wind condition are deliberated while extracting the maximum power.Wind
supply and its high penetration provide the required power to the load, but at low penetration of wind, the diesel
generator is used to meet the required
quired load. Moreover, a storage unit is connected to the system to store the energy and
to meetout the power demand. However, wind turbine generator has fluctuations in its energy production and this leads
to frequency deviation and power imbalance in the system. Suitable control strategies are essential to maintain the
power balance and frequency changes. This research work proposed an optimization approach to manage the
frequency and power control issues in the hybrid wind-diesel
wind diesel energy system using Grey Wolf Optimization. The
optimization approach is used in the research work optimizes the fuzzy P-I-D P D (proportional
(proportional–integral–derivative)
controller parameters. The proposed optimization approach is experimentally verified and compared with the
conventional PI controller, fuzzy controller, and fractional-order
fractional P-I-D
D Controller. Simulation outcomes demonstrate
that the proposed control strategy is efficient for the hybrid system with better frequency control, power balance, and
avoids system instability.
________________________________________________________________________________
____________________________________________________________________________
Keywords: Hybrid Energy system, Frequency control, Power imbalance, Grey Wolf Optimization (GWO),
Fuzzy Proportional–Integral–Derivative
Derivative (P-I-D)
(P controller.
1. Introduction pollution and cost, it is essential to integrate a

cost-effective
effective system with diesel generators for
Distributed
ibuted generation system plays a vital role
an economic environmentally friendly energy
in energy management and provides a clean
system.
tem. Among all the renewable energy
environment. To reduce the energy
sources wind energy systems are widely
requirements from a single source, alternative
integrated with diesel generators as it is
technologies such as fuel cells, photovoltaic
suitable for remote locations. Integrating wind
systems, wind turbines, diesel engines are
and diesel energy systems will be a better
introduced. In these, wind energy gets more
choice for isolated location energy
attention in recent times and the technological
requirements.Voltage
s.Voltage and frequency controllers
advancement in the present era helps to achieve
along with electronic converters are used to
better power quality by integrating wind
manage the load requirements. Under low wind
energy systems with existing systems.
conditions, load requirements are solved by
However, power quality and reliability
reliabilit are the
using diesel generators.Since the wind speed is
major factors that should be considered while
not linear and the speed mismatch intr introduces
introducing new technologies. Since a single
fluctuations in the output and it can be
power source cannot meet the necessary
eliminated using diesel generators. An
demands, and the possibility of continuous
illustration of a wind-diesel
diesel hybrid energy
supply is less. This leads to the invention of
system is depicted in figure 1.
hybrid energy systems such as wind-diesel,
wind
wind –PV,etc.,
PV,etc., along with the battery to make
sure the availability of energy sources. With
the fast controllability and quick response time,
diesel engines are mostly preferred for the
integration of energy systems.
Diesel generators are widely used in remote
Figure1 Wind-Diesel
Diesel Hybrid Energy system [3]
locations to provide power where the grid is
not available. But, due to environmental
Integrated wind turbine generators (WTG) and order, so that other than proportional, integral,
diesel generators are widely used as hybrid and derivative constants, the order of fractional
approaches. However, the dynamics of both integration and fractional derivative need to be
systems are different from each other. Hybrid considered in FPID controllers.
systems are often designed with energy storage Other than PI, PID, and FOPID, fuzzy
systems to store and distribute the generated controllers are introduced which control the
power. This energy storage system is utilized system based on fuzzy sets. A fuzzy controller
to supply during the transients for a short doesn’t require an accurate mathematical
duration. Frequency stability should be model and provides better results even for
maintained to balance the demand and supply imprecise inputs. Compared to other
in an energy system and it is applicable for controllers, fuzzy controllers perform better for
wind-diesel energy systems also. So it is complex, nonlinear systems. Based on the
essential to introduce an efficient control fuzzy sets, for the class of objects,the fuzzy
strategy while implementing the hybrid system. controller provides a smooth transition between
Various control schemes are followed in membership to non-membership functions.
conventional hybrid energy systems such as PI, Considering this advantage of fuzzy controller,
PID, fractional PID Fuzzy based models. the proposed system is developed to optimize
Proportional Integral (PI) Controller is widely the fuzzy PID controllerto minimize the power
used in industrial applications. and frequency deviations in the hybrid energy
The design and structure of the PI controller system.
are simple and cost-effective. PI controllers The Research work is structured in the
eliminate the steady-state error and forced following manner: In section 2, detailed
oscillations. However, for nonlinear and analysis of existing literaturein hybrid energy
uncertain system conditions, the PI controller systems are presented, the proposed
lags in performance. The integral model optimization approach using fuzzy PID
introduced in the controller has controloverthe controller is presented in section 3, and its
system speed and stability. Proportional experimental results, observations are
Integral Derivative (PID) Controllers are presented in section 4 followed by conclusion
employed as an industrial controller that has in the last section.
integral and differential regulator arrangements
2. Related Works
to optimize the system. PID controllers are also
termed as incremental algorithm which has Recently, researchers pay more attention to
fast response compared to PI controller. It has increase the hybrid energy system performance
different dynamic concepts such as D mode for by introducing innovative technologies. This
a quick change of controller input, I mode for section presents a literature analysis of existing
zero error by increasing the control signal, P systems to frame the motivation of the
mode to eliminate oscillations by applying proposed work. The merits and demerits of
actions inside the control system. The existing models are analyzed systematically by
derivative mode of the PID controller increase discussing the key technology.
system stability and gain. Power imbalance and frequency deviation are
The integral time constant increase the the two important factors considered in most of
controller response and due to this reasons PID the research work proposed for a wind-diesel
controller are widely adopted in hybrid energy hybrid energy system. Suitable frequency
systems. Low overshoot range and minimum control models are essential to obtain better
settling time are the major advantages of PID power quality. Similarly, variables in a hybrid
controllers. Later fractional order PID (FOPID) energy system such as dc-link capacitance,
controllers are introduced which is an kinetic energy, operating region, wind
extension of conventional PID controller which penetration, turbine size, number of turbines,
are less sensitive to parameter changes of a power reserve, power curve, and power droop
control system. The fractional-order helps the control gain, etc., should be considered in a
system to attain ISO damping easily. All the hybrid energy system design. To improve the
operations are usually represented in fractional load frequency control novel double equivalent
input disturbance controller (EID) is introduced regenerative air storage is reported in literature
in the literature [11]to fine-tune the WTG which reduces the limitation of wind turbine
results by controlling the pitch angle. Another systems in a hybrid wind-diesel energy model.
equivalent input disturbance controller is Fuel consumption minimization and
introduced to obtain balanced power output by environmental benefits are the major
adjusting the output power on the diesel
advantages of the regenerative storage model.
generator side. Compared to the conventional
PI controller, the EID controller attained better Maximum power point tracking without speed
performance in a wind-diesel energy system. sensors is achieved in research work using
Neural network and sliding mode method perturb and observe method for fixed speed
based control strategy is reported few research diesel generator and variable speed wind
works to reduce the frequency deviation and turbine. For three phase voltage source
load variance in a hybrid wind-diesel energy inverter, two control levels are introduced at a
system. It is essential that the output power of common coupling point for standalone system
diesel generators should be regulated, for that power requirements.
sliding mode control is introduced in literature. Optimization is widely adopted to increase the
Pitch angle control is smoothened using the existing system performance. The controller
sliding mode to reduce the power fluctuations. parameters in hybrid energy systems are
The neural network model is used to observe optimized using optimization algorithms.
the load variations. A combined approach Various optimization algorithms such as
attains better accuracy and stability for a hybrid particle swarm optimization (PSO), whale
energy system. The power management optimization genetic algorithm (GA),
strategy is reported in the literature utilizes the differential evaluation, other evolutionary
ANFIS model to achieve power balance in algorithms are introduced to improve the
microgrids. Along with the classification controller performance . The performance of
approach, Genetic algorithm and particle the proportional integral derivative (PID)
swarm optimization are incorporated to achieve controller is improved in literature utilized
robust power management and maximum epsilon multi-objective genetic algorithm.
PowerPoint in a hybrid wind-diesel- WTGpitch angle is optimized to minimize
photovoltaic energy system. The minimized mechanical stress and to reduce the power
cost function is the major advantage of the deviations. Moreover, the optimization
optimization approach. The dynamic multi- approach regulates the charging level of the
stage decision model is reported in the energy storage systems by decreasing the size
literature to reduce the diesel running cost and and controls the overall system frequency.
to maintain the wind-diesel system. For Nomadic people optimizer is introduced in
optimized decisions, the path nature and wind literature for a hybrid energy system as a tri-
power production are considered to reduce the objective approach by considering the
fuel cost and system instability. operators such as initial meeting, tent
Fuel consumption is one of the cost increasing distribution, family search, leadership
factors in a diesel generator. In order to reduce transition, and normal leaders periodic
the consumption, air compression and storage meeting. Life cycle cost minimization and CO2
at constant pressure approach is presented in emission reduction are the major objectives of
literature. The production cost and pollution the Nomadic people optimizer. Operational and
are reduced by introducing wind energy with financial optimization of a renewable energy
diesel generators. In conventional systems, system is reported in the literature reduces the
switch off the diesel generator on lower loads system electricity cost using mixer integer
to prevent damage is not possible. Research linear programming model. By sizing and
scheduling the components, the power supply
work overcomes the limitation and provides a
loss and maximum load capacity can be
reliable power supply to the system. Similar air obtained using the approach. Accuracy and
compression approach based energy
efficiency are the major advantages of the where∆ is the output power deviation in
optimization approach. diesel generator, ∆ is the output power
2.1 Research Motivation deviation in wind turbine generator, ∆ is the
deviated power in the energy storage unit and
Based on the above analysis the following ∆ represents the power deviation in load.
observations are presented as the motivation Figure 2 depicts the block diagram of the
behind the research work. hybrid wind-diesel energy model. The
 Power imbalance and frequency deviation generator systemon the diesel side is modeled
are mainly considered by the researchers. based on figure 1 in the frequency domain as
 Traditional models are introduced to follows
improve the performance,however; the
performances are not satisfactory for large ∆ = ∆ − ∆ (2)
systems.
( )
 Few research work utilized machine ∆ = ( )( )
∆
learning models such as ANN and ANFIS (3)
to classify the optimal parameters. though
the performance is better, it could be ∆ = ∆ (4)
further increased in optimization
approaches. where the time constants are represented as
 Further, to increase the hybrid system and speed governor gain is represented as ,
performance various meta-heuristic and the frequency deviation is represented as ∆ .
nature-inspired algorithms are introduced. Equation 4 represents the power deviation of
the diesel generator. Similarly, for wind turbine
2.2 Research Objective
generator, the power deviation is expressed as
Based on the above research findings, the
∆ = (∆ −∆ ) (5)
objective of this research work is framed to
improve the power and frequency factors of an
∆ = ∆ −∆ +
isolated hybrid wind-diesel energy system
using a Nature-inspired optimization model. ∆ 3+∆ (6)
for this the research work contributes the ( )
following ∆ =∆ (7)
 Grey wolf optimization (GWO) is proposed
to improve the power and frequency factors ∆ =∆ (8)
of an isolated hybrid energy system.
 The optimization approach is used to obtain ∆ =∆ (9)
optimal parameters of Fuzzy PID
controllers for better performance. where the fluid coupling gain in a wind turbine
3. Proposed Optimization Model is represented as , pitch control is
represented as , , and are the
The proposed optimization model is aimed to pitch responses, ∆ represents the frequency
optimize the fuzzy PID controller parameters. deviation over pitch response and frequency
Before that, a hybrid wind-diesel energy deviation is represented as ∆ . Equation 5
system is formulated using first-order block represents the WTG power deviation. The
diagram representation. the nature of wind power system is mathematically formulated as
power prediction is quite difficult so the diesel a complete system and it is expressed as
generator system is integrated with a wind
turbine generator for clean energy generation ∆ = ∆ (10)
and catering to the load demand. The power
deviation is mathematically formulated as where the gain of the power system is
∆ =∆ +∆ −∆ − represented as and represents the time
∆ (1) constant.
obtained as ⃗. The new vector positions for the

values of ⃗ ⃗ is given as
⃗ = 2 ⃗. ⃗ − ⃗ (13)
⃗ = 2. ⃗ (14)
where the vector values decrease from 2 to 0
for every iteration and position of random
vector changes into [0,1] for ⃗ and ⃗. The
position of wolves are considered as ( , ) and
Figure 2 Block diagram of Hybrid Wind Diesel the position of prey is considered as ( ’, ’).
Energy Model By regulating the vector positions ⃗ and ⃗ the
In order to optimize the controller parameters, best positions are obtained. The random values
grey wolf optimization is introduced to the are obtained using grey wolves as nodes placed
hybrid energy system. Grey wolf optimization in a random position. α and β along with δ
(GWO) is a metaheuristic optimization route the hunting operations.The hunting
algorithm formulated based on grey wolf behavior is expressed as the best solution and
hunting strategies. Generally, wolves live as a its mathematically formulated as
group and a male will be a leader which is ⃗ = ⃗. ⃗ − ⃗
termed as alpha, and female wolves take the ∝
responsibilities such as place search, hunting, ⃗ = ⃗. ⃗ − ⃗

and alert other wolves, etc., Alpha decision (15)
will be final and all the members should obey
the alpha decision. The hunting nature of ⃗ = ⃗. ⃗ − ⃗
wolves has three portions such as tracking the
⃗=
∝ −⃗ . ( ∝⃗)
prey, encircling the prey, and attacking the
prey. ⃗= −⃗ . ( ⃗)
Finding the best solution is the main objective (16)
of the wolf optimization model and the
decision is provided by alpha which is ⃗= −⃗ . ( ⃗)
considered as α, the next best solutions are
obtained by females and other members in the
group which is represented as β (beta) and δ
(delta). All the other solutions are represented
as ω(omega.) The hunting procedure will be
formulated by α,β, and δ, the other wolves ω
will follow the superior decision. wolves
identify the prey and create a circle to observe
the prey behavior and it is expressed as a
vector function as
⃗ = ⃗. ⃗ ( ) − ⃗( )
(11) Figure 3GWO algorithm-position update
⃗( + 1) = ⃗ ( ) − ⃗ ⃗ process
(12) The vector values of are expressed in
wherethe number of iteration is represented as equation (16). Figure 3 depicts the attacking
mechanism of grey wolves over the prey. the
and vector coefficients are represented as ⃗
estimated position of the prey is updated
and ⃗.. The vector position of the prey is immediately and the position of wolves is also
represented as ⃗ ( )and ⃗ is the grey wolf changed based on the prey movement. The
present position and the new position is optimization process limits its local optima
based on the new solution. If the values of The input signals for and ∆ with sampling
⃗ >1 then the process will be terminated and time 0.01s are applied to the variables such as
all the wolves will move away from the prey negative big (NB), negative small (NS), zero
and start to search other decision spaces. The (ZE), positive small (PS), positive big (PB).
pseudocode of the grey wolf optimization The output variables for the control signals are
process is summarized as follows. expressed in terms of zero (ZE), positive small
(PS), positive medium (PM), positive big (PB).
Pseudocode of the grey wolf optimization
Begin Table 1. Fuzzy control Rules
Initialize the wolves population, α, β, and δ
Obtain the best solution for the search agent Xα
Obtain the second-best solution for the search
agent Xβ
Obtain the third-best solution for the search
agent Xδ
For t=0
While (t < Max Iteration)
For each search
Update the position of search agent
End for
Update the position
Calculate the fitness function for search agents
Update Xα, Xβ, Xδ
End while
End
The above optimization approach is
incorporated to optimize the parameters of the
fuzzy PID controller in a hybrid energy system.
Based on fuzzy set theory, the logical
conditions of fuzzy are derived. Using fuzzy Figure 5 Control input variables membership
logic, the parameters of the P-I-D controller are function
improved to regulate the frequency and power
of the wind-diesel hybrid system as the Fuzzy Figure 5 depicts the membership function input
PID model. The controller has three variables which are obtained based on the IF-
processesthat include a fuzzificationprocess, THEN rule. In table 1 the relationship between
rule base and interface engine, error function and error deviationsare listed for
defuzzificationprocess. Figure 4 depicts the the proposed approach.The Grey wolf
fuzzy PID controller used in the proposed optimization approach is used in the proposed
approach.Variable errors and change in error model to obtain the Fuzzy PID controller
optimal parameters. From the analysis, the
is represented as ∆ are considered as
frequency variation obtained as an error
controller inputs in the proposed approach. PID
function should be minimized and it is obtained
parameters such as , and are the
using PID control parameters , and .
outputs calculated based on the fuzzy
controller. 4. Result and Discussion
The simulation model of the proposed hybrid
energy system for power and frequency control
is conducted to validate the efficiency of the
optimization method. A 350kW diesel
generator, 750kW wind turbine, are combined
to design the hybrid wind-diesel energy
Figure 4 Fuzzy PID controller
system, and the simulation is performed in model. Proposed model performance is

MATLAB 14.1. Figure 6 depicts the validated in two categories such as with
simulation model of the wind turbine generator optimization model and without optimization
used in the proposed model. similarly, figure 7 model to highlight the superior performance of
depicts the simulation design of the controller the proposed approach.
Figure 6 Simulation Model of Wind Turbine Generator
Figure 7 SIMULINK model of wind speed
Figure 8Frequency deviation response without optimization approach
Figure 9 Frequency deviation response with optimization approach
The frequency deviation obtained in the The performance in frequency deviation is

conventional wind-diesel
diesel hybrid energy system further improved using the optimization
without an optimization approach is depicted in approach and it is depicted in figure 9. It could
figure 8. To validate the superior performance be observed that due to the optimization
of the controller, conventionalnal PI controller, approach the range of frequency deviation is
Fuzzy Controller, fractional order PID reduced compared to figure 8.
controller is used to compare with the proposed The parameters such as peak overshoot and
model fuzzy PID controller. It could be settling time observed from the frequency
observed the proposed controller provides deviation curve are detailed in table 2. It is
minimum frequency deviation compared to observed that the proposed controller model
other controller approaches. Moreover, the provides minimum overshoot and settling time.
performance of the PI controller is poor The system stable condition is defined in terms
compared to other models as it exhibits a of rise time and If the rise time is low, the
maximum change in frequency within a short system is stable so that peak overshoot must
time. Fractional order PID controller also have obtained as stable. From the table, it
performance is near to the proposed Fuzzy PID is clear that the fuzzy-PID
PID controller exhibits
controller, however, its steady--state value is minimum settling and overshoot time for both
quite high compared to the proposed approach. conditions compared to other controllers.
Table 2 Frequency deviation analysis

Without Optimization With Optimization
S.NO CONTROLLER Peak Settling Settling
Peak Overshoot
Overshoot Time(sec) Time(sec)
1 PI controller 0.052 26 0.048 20
2 FOPID Controller 0.0375 21.6 0.042 15
3 FUZZY controller 0.0306 20 0.029 13
4 FUZZY PID Controller 0.0283 19.5 0.026 12.8
The power generated by the wind turbine case of fractional PID controller and
generators without optimization modelis conventional fuzzy controller the response is
measured and depicted
icted in figure 10. It could be similar and the difference is minimum. In that
observed that due to the efficient control fractional PID provides better performance
mechanism the power generated by the wind compared to conventional fuzzy controller.
turbine generator is steady for Fuzzy PID However,
wever, the Proposed Fuzzy PID controller
controller. In the case of PI controller, the exhibits better steady performance compared to
overshoot varies twice than Fuzzy PID other controllers.
controller that
at leads into system instability. In
Figure 10 Wind turbine generator without optimization approach

The wind turbine generated power output for When the generated power is compared with
controller models with optimization approach
a the controller without optimization results the
is depicted in figure 11. It could be observed performance of optimizat
optimization model based
that proposed fuzzy PID controller exhibits results are better. The stability of the controller
stable performance than other controllers. increases due to the optimization process.
Figure 11 wind turbine generator with optimization approach
The power output analysis performance of the controller exhibits minimum settling and
wind turbine generator is listed in table 3 in overshoot time for both conditions compared to
terms of settling time and peak overshoot time. other controllers.
It is observed that proposed fuzzy PID
Table 3 Wind turbine generator power output analysis
analysis
S.NO CONTROLLER Peak Settling Settling
Peak Overshoot
Overshoot Time(sec) Time(sec)
1 PI controller 0.705 27.5 0.6 19
2 FOPID Controller 0.5673 20.4 0.51 14.2
3 FUZZY controller 0.5261 19.5 0.36 13.8
4 FUZZY PID Controller 0.5215 19.2 0.34 11
performance. In the case of the PI controller,

The power generated in the hybrid energy
the sudden increase in the output may damage
system using wind-diesel
diesel generator is depicted
the system. In case of fuzzy controller and
in figure 12 without optimization and figure 13
Fractional PID controller the performance is
depicts the performance analysis of generated
better compared to PI controller however, it is
power with optimization approach. compared
less than the proposed Fuzzy PID controller.
to other controllers the proposed Fuzzy PID
controller exhibits stable and better
Figure 12 Wind-diesel
diesel generator without optimization approach - Generated power analysis
Figure 12 Wind-diesel
diesel generator with optimization approach-
ap Generated power analysis
Table 4 depicts the details of power generated controllers are observed with optimization and
by hybrid wind-diesel energy system in terms without optimization model and it is visible
of settling time and peak overshoot time. The that the proposed fuzzy PID model exhibits
performance of proposed controller and other superior performance in all aspects.
Table 4 Power output analysis for wind-diesel generator
S.NO CONTROLLER Peak Settling Peak Settling
Overshoot Time(sec) Overshoot Time(sec)
1 PI controller 1.11 25 1.11 18
2 FOPID Controller 0.1315 20.25 0.12 12
3 FUZZY controller 0.1011 19.95 0.1011 11.6
4 FUZZY PID Controller 0.0701 17.45 0.06 10.2
The optimizer plot for the grey wolf
optimization module is depicted in figure 13.
The functions are presented on both sides and
the optimal solution is obtained for the given
functions. In the proposed approach, PID
controller parameters are considered as
objective functions and the optimal solution is
obtained to improve the parameters.
Figure 14 Obtained best score
Table 5 Objective function

Control schemes Objective
function
PI controller 0.8571
Fractional order PID 1.061
controller
Fuzzy controller 5.887
Fuzzy PID controller 5.489
The best score obtained for the given objective
Figure 13 Grey wolf optimization module is depicted in figure 14. It is observed that the
proposed fuzzy PID controller attains better
score than other models due to the optimal framed in the research model. to improve the
parameters. though the score between fuzzy performance of controllers, optimal parameters
and fuzzy PID are close there is a huge are identified in the proposed approach using
variation in other performance analyses. Table grey wolf optimization. To validate the
5 depicts the objective function values for the proposed approach performances, simulation
different controller approaches. It is observed design is obtained and analyzed for the
from the table the objective function of the frequency deviation, the power generated in
proposed Fuzzy PID controller and the fuzzy wind turbines, and power generated in the
controller havea maximum objective function wind-diesel energy system in terms of settling
value that indicates the proposed approach time and peak overshoot time. Conventional
covers the maximum extent to obtain better controller models such as PI controller, fuzzy
performance. From that, it is concluded that the controller, and fractional order PID controller
fuzzy PID controller attains better performance are used to compare with the proposed fuzzy
than the conventional fuzzy controller. PID controller. Proposed approach simulation
results are obtained for two categories as with
5. Conclusion
optimization and without optimization model
An optimized power and frequency control for and it is observed that for both cases fuzzy PID
a hybrid wind-diesel energy system is controller performs better. This research could
presented in this research work. The necessity be further developed by introducing a hybrid
of frequency control and power balance optimization approach to increasesystem
strategies in isolated systems are identified and accuracy and efficiency.
References
1. Abbas Q. Mohammed; Kassim A. Al- management – Experimental

Anbarri; Rafid M. Hannun (2020), investigation” Energy, vol. 130, pp. 38-47.
“Optimal Combination and Sizing of a 6. Félix Dubuisson; MiloudRezkallah;
Stand–Alone Hybrid Energy System Using Ambrish Chandra; MaaroufSaad; Marco
a Nomadic People Optimizer”, IEEE Tremblay; Hussein Ibrahim (2019),
Access, vol. 8, pp. 200518-200540. “Control of Hybrid Wind–Diesel
2. AsitMohanty, MeeraViswavandya, Standalone Microgrid for Water Treatment
PragyanParamita (2015), “Optimisation System Application” IEEE Transactions
and Improvement of Voltage stability in a on Industry Applications, vol. 55, no. 6,
Standalone Wind-diesel-micro Hydro pp. 6499-6507.
Hybrid System” Procedia Technology, 7. GrzegorzIwanski, ŁukaszBigorajski,
vol.21, pp.332-337. WlodzimierzKoczara (2018), “Speed
3. Baseem Khan; Pawan Singh (2017), control with incremental algorithm of
“Selecting a Meta-Heuristic Technique for minimum fuel consumption tracking for
Smart Micro-Grid Optimization Problem: variable speed diesel generator”, Energy
A Comprehensive Analysis”, IEEE Conversion and Management, vol. 161, pp.
Access, vol. 5, pp. 13951-13977. 182-192.
4. ChunSheng Wang; Jiamin Li; Yukun Hu 8. Haixin Wang; Junyou Yang; Zhe Chen;
(2019), “Frequency Control of Isolated Weichun Ge; Yunlu Li; Yiming Ma; Jian
Wind-Diesel Microgrid Power System by Dong; Martin OnyekaOkoye; Lijian Yang
Double Equivalent-Input-Disturbance (2019), “Analysis and Suppression for
Controllers”, IEEE Access, vol. 7, pp. Frequency Oscillation in a Wind-Diesel
105617-105626. System”, IEEE Access, vol. 7, pp. 22818-
5. DjohraSahebKoussa, M. Koussa, S. 22828.
Bellarbi (2017), “Hybrid diesel-wind 9. Hesham M. Fekry; Azza Ahmed
system with battery storage operating in Eldesouky; Ahmed M. Kassem; Almoataz
standalone mode: Control and energy Y. Abdelaziz (2020), “Power Management
Strategy Based on Adaptive Neuro Fuzzy
Inference System for AC Microgrid”, Performance Using Multi-Objective

IEEE Access, vol. 8, pp. 192087-192100. Optimization” IEEE Access, vol. 5, pp.
10. Ibrahim.H., Lefebvre.J, Methot. J.F, 6212-6224, 2017.
Deschênes, Jean-Sébastien (2011) “No- 17. Neerparaj Rai, Bijay Rai (2018), “Control
storage wind-diesel system: Mechanical of fuzzy logic based PV-battery hybrid
modeling based on power flow models” system for stand-alone DC applications”
2011 IEEE Electrical Power and Energy Journal of Electrical Systems and
Conference, EPEC 2011. pp. 17-22. Information Technology, vol. 5, no. 2, pp.
11. Inés Tejado, Blas M. Vinagre, José Emilio 135-143.
Traver, Javier Prieto-Arranz, Cristina 18. RaffaelEngleitner; AdemirNied; Mariana
Nuevo-Gallardo (2019), “Back to Basics: Santos Matos Cavalca; Jean Patric da
Meaning of the Parameters of Fractional Costa (2018), “Dynamic Analysis of Small
Order PID Controllers” Mathematics, vol. Wind Turbines Frequency Support
7, no. 530, pp. 1-16. Capability in a Low-Power Wind-Diesel
12. Leong Kit Gan, Jonathan K. H. Shek, Microgrid”, IEEE Transactions on
Markus A. Mueller (2016), “Optimised Industry Applications, vol. 54, no. 1, pp.
operation of an off-grid hybrid wind- 102-111.
diesel-battery system using genetic 19. Saad.Y, Younes.R, Abboudi.S, Ilinca.A
algorithm”, Energy Conversion and (2018), “Hydro-pneumatic storage for
Management, vol. 126, pp.446-462. wind-diesel electricity generation in
13. Mahmood Ul Hassan, Muhammad remote sites” Applied Energy, vol. 231,
Humayun, RizwanUllah, Baoquan Liu, pp. 1159-1178.
Zhuo Fang (2018), “Control strategy of 20. Sebastian C. Manchester, Lukas G. Swan,
hybrid energy storage system in diesel Dominic Groulx (2015), “Regenerative air
generator based isolated AC micro-grids”, energy storage for remote wind–diesel
Journal of Electrical Systems and micro-grid communities” Applied Energy,
Technology, vol. 5, no. 3, pp. 964-976. vol.137, pp.490-500.
14. MiloudRezkallah; Sanjeev Singh; Ambrish
Chandra; Bhim Singh; Marco Tremblay; 21. Shailendra Kumar Tiwari; Bhim Singh;
MaaroufSaad; Hua Geng (2019), Puneet Kumar Goel (2019), “Control of
“Comprehensive Controller Wind–Diesel Hybrid System with BESS
Implementation for Wind-PV-Diesel for Optimal Operation”, IEEE
Based Standalone Microgrid”, IEEE Transactions on Industry Applications,
Transactions on Industry Applications, vol. 55, no. 2, pp. 1863-1872.
vol. 55, no. 5, pp. 5416-5428. 22. TuTu, Gobinath P. Rajarathnam, Anthony
15. Minghan Yuan; Yang Fu; Yang Mi; M. Vassallo (2019), “Optimization of a
Zhenkun Li; Chengshan Wang (2018), stand-alone photovoltaic–wind–diesel–
“The Coordinated Control of Wind-Diesel battery system with multi-layered demand
Hybrid Micro-Grid Based on Sliding scheduling” Renewable Energy, vol. 131,
Mode Method and Load Estimation”, pp. 333-347.
IEEE Access, vol. 6, pp. 76867-76875. 23. Yu Hu, Juan M. Morales, Pablo Solana
16. Mohammed E. Lotfy; TomonobuSenjyu; (2015), “Dynamic multi-stage dispatch of
Mohamed A. Farahat; Amal F. Abdel- isolated wind–diesel power systems”
Gawad; Atsushi Yona (2017), Energy Conversion and Management,
“Enhancement of a Small Power System vol.103, pp. 605-615.
ROLE OF INFORMATION TECHNOLOGY IN ENERGY HEALING TECHNIQUES - A

META ANALYSIS
Jeo J. A1,2*, Molly J.1 and Krishna P. K2
1
Kristu Jayanti College (Autonomous) Bengaluru
2
Srinivas University, Mangaluru
*jeojoy@kristujayanti.com
_______________________________________________________________________________________
ABSTRACT
Human body behaves like a rechargeable battery which utilizes the energy on the earth and the entire universe for
recharging itself. The mother earth and the entire universe help and provide us everything needed for the spiritual,
mental and bodily healing. Besides human beings have a constant flow of energy from the feet to the minds. The
Energy Healing Techniques believe that this flow of energy can heal the body by itself in a natural way. The paradigm
shift from the traditional medicine system which reveals us about the meridian system and the energy chakra vortexes
to the modern science all tell us about energy system and be indebted to the basic principle that every substance is
produced of energy. Modern era medical practitioners have recognized the benefits of energy healing and have started
implementing it as one of the supplements to the modern medical treatment. This review provides light into the
classification of energy healing techniques, global and national impact of Energy Healing Techniques, the various
conventional and modern approaches for the measurement of biofield energy and how Information Technology is
involved in the various techniques. Energy healing techniques and the use of computers and other information
technology in the therapeutic modalities were identified and analyzed using databases from different sources.
Information was extracted from the topic related books and journals in addition to the relevant websites. This review
summarizes various conventional and modern approaches for measurement of biofield energy and the use of computers
and information technology for the same. The structural analysis infers that extended usage of computerized imaging
and software analysis is the need of the hour for scientific evidences of the effectiveness of these therapies.
________________________________________________________________________________
Keywords: Energy Healing, Biofield Energy, Role of Information Technology.
Introduction The human brain sends signals which ranges in

the frequencies between 20 and 100 Hz to the
Human beings do not consist of only physical
muscles, and energy patterns can be observed
matter. The human body can be considered as
in the concurrent diagnosis of the brain and
an embodiment of wider spectrum of energy
muscles. Ganglions along the spine is
fields which vibrates at different levels of
observed to emit electromagnetic energy
refinement. It can be manifested that our
frequencies. The root cause of electromagnetic
energy fields encompass whatever that has
energy in the nerve centers is basically due to
occurred in our life time. If this energy field is
the electromagnetic nerve impulses and
not in a balanced state then it may lead to
biochemical processes through hormones like
diseases.
oxytocin. The centers of energy in the human
Human nervous system has specialized
beings are of considerable size spanning across
subsystem called autonomic nervous system
several centimeters . Another matter of
which controls activity like breathing,
importance is that the electromagnetic energy
sweating, digestion, and even the beating of
radiated by the tissues cannot be local. The
our heart and no need for us to give these
energy field radiated from the human heart can
functions a basic thought. Our memory,
be measured several feet away from the body.
mobility and senses can function normally
Frequencies below 100 Hertz are used in brain
from the brain energy. Psychological and
research studies using electroencephalogram
emotional energies are transmitted in the form
(EEG) and those below 200 Hertz are used in
of electrical impulses along the neural and
electromyography (EMG). Human beings are
meridian pathways of the human body. These
considered as bodies that are constantly
energy transmissions in association with the
vibrating in unique electromagnetic waves
vibrations of atoms and molecules produce
(healthy human Bio field is of frequency 62-68
electromagnetic fields in the form of Human
MHz) .
Bio Field.
The term energy was first depicted in ancient Generally Practiced Energy Healing
Indian texts like Upanishads and in Traditional Techniques
Chinese Medicine almost five thousand years
Techniques based on meridian system:
earlier. Energy healing is the healing Acupuncture/Acupressure are ancient
methodology which existed in the past Chinese therapies and they work for the
which focuses on the restoration of human balancing of chi/ki/qi energy flow in the
energy body as well as maintains the right meridian system of the body. Emotional
balance of the energy flow in the different Freedom Technique (EFT) which uses
planes such as human body, mind and the mind-body medicine in combination with
soul. needle-free acupuncture. The two skills used
The physical illness, mental illness and in EFT are tapping locations and the framing of
even financial issues to an extent are reminder phrases regarding the problem which
considered as issues related to the energy is addressed. Donna Eden’s Healing
blockages under the Energy Healing Technique is associated with meridians,
paradigm. According to this it identifies chakras, aura and the other five energy
the basic cause of any illness as the systems namely the basic grid, the celtic
blockage or imbalance of the energy flow weave, the five rhythms, triple warmer, and the
in the human body. When this blockage or radiant circuits. In this technique the body
imbalance in the energy flow is fixed, energies weave themselves into celtic
curing takes place inherently. This is the weave which are present at the cellular as
basic principle of working of these well as beyond the aura level and this celtic
techniques. The efficiency of these weave is worked upon for the restoration of
techniques depends on Healer-Client health.
relationship which involves attributes such
as collaboration, trust, compatibility, Energy-Focused Bodywork Techniques:
communication and trust. Therapies like Lomi, Shiatsu, Trager and
Craniosacral Therapy (a light-touch manual
Energy Healing Techniques therapy that works at the blockages in the
The techniques which are considered as a craniosacral system) are examples of these
subset of complementary and alternative which work intentionally on this subtle energy.
medicine rooted on the belief that a vital
Distance Healing Intention Therapies:
energy flows through the human body whose Distance Healing Intention (DHI) which is
goal is to balance the energy flow in the human defined as an act directed towards the
body. improvement of health of a person at a far off
HER (Healing Energy Research) is widely distance from the healer; Johrei a Japanese
used to define areas which describes healing technique which comprises of energy
intentional influence of living beings on channeling and light massage; Reconnective
one another without the usage of physical Healing which is a spectrum healing technique
medium. We can categorize the Energy involving energy, light and information.
Healing techniques as follows:
 Traditional Eastern Asian Techniques Other Therapies: Healing Touch (HT) which
such as Reiki and qigong works with various energy fields of the body
 Professional Traditions practiced in the and chakras; Intuitive Healing which uses
West which is often practiced by nurses intuition for the healing process; Polarity
like Therapeutic Touch and Healing Therapy (PT) which uses principles of
Touch. attraction, repulsion and neutrality; Pranic
 Eastern European bioenergy traditions. Healing which works on the basic principle of
 Contemporary Metaphysical Traditions cleansing or eradicating dirty energy from the
which have eclectic backgrounds. body and substituting it with prana; Quantum
Touch which focuses on practitioner’s
breathwork, intentions, energy sensations and
touch; Reiki where practitioner’s channel
universal energy through their hands into the The regulations to be followed by homeopathy
patient’s body; Qigong which focuses on practitioners were updated in 2014 and the
regaining the balance of thee body through updated list of registered Traditional Medicine
movements of the body in coordination with ™ practitioners was made in 2016.
breathing along with meditation; Reflexology,
a Chinese based intense massage therapy
which improves overall improvement of the
lymph flow; Chakra Healing which works on
proper alignment and synchronous
ronous working of
the chakras which will allow a steady flow of
energy throughout the body ; Shamanic
Healing where Shaman is the one who makes
in roads to unordinary reality in a changed state
of consciousness. Global Analysis Report of Energy Healing
Techniques
Traditional and Complementary Out of 179 WHO member countries 98 of them
Medicine(T&CM) – An Indian Perspective have framed policies at the national level on
AYUSH which monitors the departments of Traditional and Complementary/Alternative
Ayurveda, Yoga, Unani, Siddha and Medicine (T&CM)3. Eighty
Eighty-eight percentages
Homeopathy was formed in 2014 under the of countries among 170 countries that have
Ministry of Health, India, include various membership in WHO have acknowledged tthe
expert committees and councils for monitoring usage of T&CM .
the respective disciplines
sciplines including the T&CM.
T&CM Usage South-East

East Asian Region (11 Countries)
14% 6% European Region (53 Countries)
11% 27% African Region (47 Countries)

American Region (35 Countries)
18%
Eastern Mediterranean Region (21 Countries)
24% Western Pacific Region (27 Countries)
The usage of different types of T&CM for 133 countries is as given below:
T&CM Types Usage (N=133)

120
100
80
60
40
20
0
Homeopathy Indigenous Accupuncture Herbal Ayurvedic Unani Traditional Naturopathy Others Osteopathy Chiropractic
Traditional Medicines Medicine Medicine Chinese (includes EH
Medicine Medicine Techniques)
From the graph given above it is evident that acupuncture point conductivity measurements,
among the various types of T&CM techniques and (3) bio photon measurements.
the Energy Healing Techniques are least used.
(1) High-Voltage Electrophotography
Forty percentages (Seventy-Five countries)
have reported to have National Research The gas discharge visualization (GDV) camera,
Institute for T&CM. developed by the Dr. Korotkov in 1948 is the
best among the contemporary high-voltage
Bio Field Measurement Devices in Energy electrophotography technique which is based
Healing on the Kirlian effect and uses pulses (10-
These are physical instruments used for microsecond) of high-frequency (1024 Hz) and
measuring bio field interactions. They fall high voltage electricity (10-15 kV). A charge-
under two broad categories: Diagnostic coupled detector (CCD) is used here in order to
Devices which measure bio field properties and detect the pattern of photons emitted from each
Therapeutic Devices in order to manipulate fingertip. This information is then analyzed in
biofield interactions. These devices work the computer using GDV software for various
under different modalities such as parameters of the emission patterns including
Electromagnetic Field (EMF)-light, heat, non- area, intensity, density, and factuality, as well
thermal; electrical current, vibration and sound; as details of various sectors of the fingertip
physical and mechanical; intentionality and no patterns that purportedly relate to the
locality; gas and plasma. bioenergetics of specific organs and organ
systems.
Conventional Measurement of Human Bio
Field (2) Acupuncture Point Conductivity
Conventional medical science used Measurements
Electrocardiogram (ECG) developed in 1887 Here assessment of electrical conductivity of
for tracking the electrical signals of heart and the skin is done through the acupuncture
Electroencephalogram (EEG) developed in meridian system for the purpose of providing
1875 for recording the electrical signals of the information about the energy flow related to
brain. For tracing the magnetic field the health of the body. This technique is
measurements of the heart and brain magneto otherwise known as electro dermal screening
cardiogram (MCG) and magneto (EDS), electro dermal testing (EDT), and
encephalogram (MEG) are used respectively. electro acupuncture (EAV). The primary usage
In order to capture the very low human of these devices is in allergy testing.
magnetic fields specialized equipment’s like
(3) Bio Photon Measurements
Super-conducting Quantum Interference
Device (SQUID) were used. Medical scans, Measurements of extremely weak light
such as Functional Magnetic Resonance emission from the body which is in the visible
Imaging (FMRI) and Positron Emission range from 400 to 720 nm in wavelength, is
Tomography (PET) were also used for tracking one of the techniques used for checking the
the electromagnetic activity. Galvanic Skin radiant non-thermal human bio field. For the
Response (GSR) was used as a major measurement of bio photons the various
technology for lie detection. Medical imaging methodologies used include: (1)
systems such as thermography uses the infrared photomultiplier tubes, which is used for
camera for capturing the emissions (which registering the photon count over time; (2) a
were infrared in nature) from the human body spectral analysis system, which uses a set of
on the order of 100 watts. cut-off optical filters to determine the
wavelength characteristics of the emitted light;
Modern Approaches for the Measurement (3) and a two-dimensional system of sensitive
of Bio Field photon-counting devices, including arrays of
The modern approaches for the measurement cooled photomultipliers and CCDs that
of bio field falls under three categories: (1) produce bio photon images.
high-voltage electrophotography, (2)
Information Technology in Energy Healing process which can create full colour bioenergy
charts of objects, animals, plants and even
Poly contrast Interference Photography (PIP)
ambient energy fields. In this technique the
developed in 1980’s by Harry Oldfield is
practitioner holds the RFI™ Digital Frequency
another imaging system which does the
Counter with a specially tuned antenna in his
analysis of the light reflected from the object
hand, and place the antenna in each part of the
being scanned with the help of a computer
bioenergy field where you want to detect the
program. The light would be interfered by the
colour. As the sensor tells you the objective
energy field both when the incident ray
frequency in standard scientific units, you
travelled towards the object and when the
simply enter that number into
reflected ray bounced off the object. Computer
the RFI™ software program. Once the
program is used for the analysis of different
measurement of all the regions of the Aura is
light intensities being reflected from the person
done the software can automatically generate
or object being scanned. The software
full colour images of the Aura, the Brain, or
associated with this technique can identify
both. Distance Healing can be made possible
distinctly the various patterns of light as they
through various modes starting from the
are already calibrated in the machine before in
simplest of all, that is, by email. Webinars and
hand. The different pattern of colours signifies
Internet Conferences shall also facilitate
a distinct inference and is used to portray the
Distance Healing Practices. Another pivotal
positive vibrations and imbalances in the bio-
example for the role of Information Systems is
field.
in the Energy Healing is the aura scanning
Digital Aura Scanner (DAS) is an imaging
reports generated by GDV camera while
system based on the thermal energy was
scanning fingers of both left and right hand of
developed in 2007. This imaging system uses
the person. The report generated gives us an
an ordinary digital camera and a computer.
idea about the energy flow in the body, chakra
Software which is customised to take the
alignment and predicts the malfunctioning of
scanning. This software uses photon
various organs within the body. These reports
interference to scan in and around the body.
help the healer to take a look at the chakra
This is a computer programming innovation
which allots a number to each specific analysis before and after the healing process.
The GDV camera can be associated with the
frequency of photons and re-codes each
GDV Sputnik, which uses sensors to evaluate
number into the visible light range. After
the energy field of the environment around.
processing, the image is displayed on the
The auxiliary device Biocor can be used for
computer screen and energy intensity variances
correcting the energy state with the help of
can be noted. We don't see the difference in
very high frequencies.
energy absorption with our own eyes because
we use heuristics that generalize colours, Conclusion
patterns and shapes so that we can better The Energy Healing Techniques owes its
distinguish borders and objects. Every colour existence way back in the past and is still
has a diverse frequency and, in this system, continued to exist as one of the Alternative
they represent a different stage and condition. Medical practices all over the world. Even
Here pictures are taken for the subject using though we have different practices globally,
full-spectrum lighting against a white most of these practices are focused on the basic
background. These pictures are then processed phenomenon of energy accumulation and flow.
by the computer so that we are able to see the As the WHO report states the usage of these
energy field as a series of patterns and techniques have not encompassed the whole
symmetry showing the infected areas as world as on date. The lack of scientific devices
distorted and darker and help us to understand for quantifying the bio-field is another
true root cause of the disease and enable us to drawback in the research advancement in this
visualise all seven chakras or glands. area. The impact of these therapies on the
Resonant Field Imaging (RFI) is an healer and the person who is healed can be
electromagnetic measurement and imaging scientifically explored in a wider horizon.
Even though we have enormous advancements person and perceive energy blocks, leakages,
in Information Technology less is known about and ranges of inflammation and weakening in
the usage of these technologies in the Energy the body parts. This study is meant to serve as
Healing practices. Wider scientific research a pilot for future research. Though difficult, it
can be carried out on the usage of information is imperative from a public health standpoint to
technology in all the areas of Energy Healing shift through the concepts of Complementary
practices and more tools should come into and Alternative Medicine (CAM) and learn
existence which will prosper the usage of these how alone or in combination they can improve
techniques all over the world. Taking Photos upon our current standard of health care.
or Scanning the clients facilitate the specialist Software development related to the domain of
to assess the field of energy for better analyses Energy Healing engages a lot of researchers for
of their health-related problems. When used, the development of software system. More and
the operator and client can uncover the root more National Research Institutes focusing on
disharmonies of their disease and DAS this area should come up all over the world.
therefore enables effective monitoring of a Assumptions that most of these therapies do
client's ongoing condition. Analyzing at the not have any side effects are to be proven
DAS image, we can assess the energy field of a scientifically.
References
8.
1. Ann L. Baldwin and Natalie L. Trent, “An Hammerschlag R, Jain S, Baldwin AL, et
Integrative Review of Scientiﬁc Evidence al. Biofield research: a roundtable
for Reconnective Healing”, The Journal of discussion of scientific and methodological
Alternative and Complementary Medicine”, issues. J Alternat Complement Med. 2012;
Volume 23, Number 8, 2017, pp. 590–598 18: 1081–1086.
2. Ann Marie Chiasson, “Energy Healing: 9. https://www.who.int/traditional-
The Essentials of Self-Care”, United States complementary-integrative-
of America, 2013 medicine/WhoGlobalReportOnTraditional
3. Aoyagi Y, Stein RB, Branner A, et al. AndComplementaryMedicine2019.pdf?ua=
Capabilities of a penetrating microelectrode 1
array for recording single units in dorsal 10. Jens Rowold, PhD, Diploma and Paul D
root ganglia of the cat. J Neuroscience Hewson, Biofield Frequency Bands—
Methods. 2003; 128: 9–20. Definitions and Group Differences, Global
4. B. Shanmuga Priya, R. Rajesh, Advances in Health and Medicine 2020,
“Understanding Abnormal Energy Levels in Volume 9: 1–10
Aura Images”, ICGST AIML-11 11. John Upledger, Tad Wanveer “Helping the
Conference, April 2011 Brain Drain: How Craniosacral Therapy
5. David Muehsam, Gaétan Chevalier, Tiffany Aids ADD/ADHD”, 2009
Barsotti, Blake T. Gurfein, “An overview of 12. Kuman M, J Altern, “Measuring Reiki
biofield devices”, Biofield Science and Healing - Mystery, Placebo or Real Energy
Healing: Toward a Transdisciplinary Healing”, HSOA Journal of Alternative,
Approach, 2015, pp. 42-51 Complementary & Integrative Medicine,
6. Dean Radin, Marilyn Schlitz, Christopher Volume 3, Issue 1, 2017
Baur, “Distant Healing Intention therapies: 13. McCraty R, Atkinson M, Tomasino D,
An overview of the scientific evidence”, Bradley RT. The coherent heart. Heart-
Biofield Science and Healing: Toward a brain interactions, psychophysiological
Transdisciplinary Approach, Research Gate, coherence, and the emergence of system-
November 2015 wide order. Integr Rev. 2009; 5: 10–115.
7. Fridlund AJ, Cacioppo JT. Guidelines for 14. Mima T, Hallett M. Corticomuscular
human electromyographic research. coherence: a review. J Clin
Psychophysiology. 1986; 23:567–589. Neurophysiology. 1999; 16: 501.
15. Moreno-L_opez Y, Mart_ınez-Lorenzana Health by Measuring the Biofield Aura: An

G, Cond_es-Lara M, Rojas-Piloni G. Overview”, International Journal of
Identification of oxytocin receptor in the Applied Engineering Research, ISSN 0973-
dorsal horn and nociceptive dorsal root 4562 Vol. 10 No.35 (2015)
ganglion neurons. Neuropeptides. 2013; 47: 23. Wayne B. Jonas and Cindy C. Crawford,
117–123. “Science and Spiritual Healing: A Critical
16. Motoyama H. Theories of the Chakras: Review of Spiritual Healing, Energy
Bridge to Higher Consciousness. Wheaton, Medicine, and Intentionality”, Alternative
IL: Theosophical Publishing House; 1988. Therapies in Health and Medicine, 2003,
17. P. H. Canter, L. B. Brown, C. Greaves and pg. 56
E. Ernst, “Johrei Family Healing: A Pilot 24. https://timesofindia.indiatimes.com/life-
Study”, Access Publication, July 2006 style/health-fitness/home-remedies/5-most-
18. Rubik B(2016), Measurement of the Human effective-energy-healing-techniques-and-
Biofield and Other Energetic Instruments how-they-work/articleshow/70698807.cms
19. Rubik B. The biofield hypothesis: its 25. https://www.the-energy-healing-
biophysical basis and role in medicine. J site.com/types-of-energy-healing.html
Alternat Complement Med. 2002; 8:703–717. 26. https://www.spiritualresearchfoundation.or
20. Sara L. Warber, Deogracia Cornelio, g/about-us/methodology-of-
Jeremy Straughn, and Gaia Kile, “Biofield research/polycontrast-interference-
Energy Healing from the Inside”, The photography-pip
Journal of Alternative and Complementary 27. https://www.barbarabeckerhealing.com/lon
Medicine, Volume 10, Number 6, 2004, pp. g-distance-healing-part-iii-by-computer-
1107–1113 technology/
21. Sara L. Warber, Rosalyn L. Bruyere, Ken 28. http://jmshah.com/aura-report-biowell-
Weintrub, Paul Dieppe, “A consideration of gdv.html
the Perspectives of Healing Practitioners on
29. http://gdvsoftware.com/gdv-software-
Research Into Energy Healing”, Global
history
Adv Health Med. 2015, pp. 72-78.
22. Shreya Prakash, Anindita Roy Chowdhury,
Anshu Gupta, “Monitoring the Human
STRATEGIZING THE HIGHER EDUCATION IN INDIA FOR FUTURE –A

TECHNOLOGY-DRIVEN APPROACH IN EDUCATION
K. Anusha1, D. Praveenadevi2 and S. Poojitha1

1
KL Business School, Koneru Lakshmai Education Foundation, Vaddeswaram, Guntur, AP, India
2
K..S. Rangaswamy College of Arts and Science, Namakkal District, Tamilnadu
kanagalanusha@gmail.com
_______________________________________________________________________________________
ABSTRACT
Background/ Objectives: To understand the role of Indian higher education in promotion of social cohesion, civic
behavior and nation building and; to design for a transforming model in strengthening the higher education with
formulation of world class generic strategies. Methods: A model is constructed to assess the role of higher educational
institutions in pandemic. The Institutions are categorized into four categories: mass customization, Innovation, mass
production and continuous improvement considering the parameters the product change and process change.
Discussion: The higher educational institutions help in strengthening of social capital. The emergence of quality and
innovation in their approach is a need for higher educational institutions to formulate world class generic strategies
and evaluate their competiveness and readiness.
________________________________________________________________________________
Keywords: Higher Education, Information Technology, Quality, Pandemic, Strategy
Introduction under the act of parliament without

accreditation the government notes the
The Indian higher education system is the
institution to be fake. Accreditation is overseen
third-largest in the world. India is the second-
by an autonomous institution like All India
largest in the e-learning market after the United
Council for Technical Education (AICTE),
States. The UGC (University grant
Indian Council of Agricultural Research
commission) in India is the main governing
(ICAR), National Assessment and
body to enforce the standards. The Indian
Accreditation Council (NAAC), Pharmacy
higher education system is adding millions of
students every year. India has a total of 967 Council of India(PCI), Bar Council of India
(BCI), etc. established by UGC. Universities
universities in the country as of 2020, among
are ranked to measure Quality on the range of
967 universities: 54 are central universities,
parameters related to research, reputations, and
418 state universities, 125 deemed to be
teaching. The National Institutional Ranking
universities, and 70 private universities. In
Framework is a mechanism to administrate the
India, the universities are monitored by an apex
quality for universities and institutions.
body. The universities are indirectly controlled
by the ministry of education and funded jointly Review of Literature
by the state government apart from state Indian Education Institutions rank second in
universities. They are premium research student enrollment to higher education and are
institutes like the National Institute of competitively advantageous as the country has
Technology (NIT) Indian Institute of a large English-speaking population. India
Technology (IIT) Indian Institute of ranks 28 out of 88 countries in the English
Information technology (IIIT) Indian Institute proficiency Index. In India, UGC granted the
of Science (IISc) across the country in science NAAC A+ Universities to offer Online
and technical education. The enrollment is Learning. Most Millennial (aged 22-35) are
increasing correspondingly and the quality into online certification and in re-skilling for
dimension is being mandated for institutions career advancement. Digital Landscape in
across the country. The private sector is strong India, internet usage and penetration is rapidly
in Indian higher education. Institutes are increasing to reach 550 million population
sourcing international students around the Content is important in the education industry.
world to enroll in master's programs. The The content through e-learning is through self-
universities to function have to be accreditation paced, live e-learning. Since the lockdown in
India, about 300 million students are dependent The Need for IT Integration to Higher
on education online. The major segments Education Institutions
accelerated at a fast pace online in each Information Technology assesses how
segment: Primary and Secondary -60%, Test Information Technology has facilitated the
Preparation -64%, Up skilling and Online growth in interactive learning and what has
Certification- 38%, Higher Education -41%, been its impact in the higher education scenario
and Language and Casual Leaning -42% in the country. The Outbreak of COVID-19
CAGR (KPMG,2019). The Indian Education caused a sudden closure of campuses to
system needs to be educated to use information prevent transmission and ensure social
technology; otherwise, purchase and transfer of distancing. The economic impact is highly
technology. The pros of traditional classroom visible. The educators had to focus on online
teaching and ICT tools should be integrated learning systems. The need for technology
into a single system. The drivers for online integration was important for the institutions
education in India include internet penetration, by overcoming the limitations as accessibility
government initiatives, and affordability. and affordability for many students. The
Massive Open Online Courses (MOOCs) offer education climate changed rapidly with the use
the flexibility to take select courses online via of e-learning and virtual learning. The
the web. They are interactive, provide Pandemic turned out to be ideal to experiment
community interactions among students and and deploy new technology to make education
professors also provide with the assignments, delivery possible and meaningful.
quizzes and emerged to be a popular mode of
learning.
No of Students Enrolled in Distance/ Online (Degree/Diploma) Programs in

India (in Lakhs)
70
62.64
58.67
60
51.01
50
40.31 41.45 Distance Education Enrollments
40
32.25 Online Education
30 Enrollments(Degree)
Online Education
20 16.12 Enrollments(Diploma)
10
3.22
0.92 0.95 0.97 0 0.99
0
2018 2019 2020 2021 2022
Source: Schoolguru-valuvox analysis, January 2019

Higher Education Institutions Strategies assess where are we now? Also where we want
for Information Age to be (Vision)
How will we get there? (Strategy)
The future development of educational
In a business environment, Institutions are now
institutions is wholly dependent on strategic
experiencing rapid and continuous change,
thinking i.e. 'Predicting the future. Strategic
which can be identified in terms of product
thinking answers the questions like are the
change and terms of process change. Product
institutions familiar with SWOT analysis to
change characterizes the demand for new
courses or new services. Institutions change teaching and the integration of technologies to
their courses because of competitive moves; deliver the outcome or services. Process
customers' preferences are the emergence in broadly refers to faculty, accreditation systems
the nation. Process change refers to altering the technologies, and procedures used to develop
procedures like the methodologies used in the institution.
The matrix is a classified change to be either Online learning portals are online knowledge-
stable or dynamic. Stable change is slow and sharing tools. A learning tool helps to trace the
generally predictable. Dynamic change is information needed quickly. Below are
revolutionary, rapid, and generally mentioned with the top learning portals.
unpredictable. The two types of changes Bharat Skills is a repository for skill-based
provide four possible combinations
nations of change courses. Have e-learning
learning videos available for
conditions that can assess an organization to trades like electricians, carpenters, plumbers,
understand its competitive position, etc.
infrastructure, and IT strategies and processes Skill India is a multilingual portal for about
appropriately. 400 courses in 10 different regional languages.
NAASCOM-FutureFuture Skills is an industry
industry-driven
Dynamic Product and Process Change:
learning ecosystem for the students and
Innovation Strategy
professionals to re-skill.
The Institution focusing on Innovation
Innovati strategy Skills build provides courses in the latest
frequently create a new course, constantly technology like cyber security, Artificial
innovates the process required to develop and Intelligence along with soft skills.
produce the new methodologies into the new TCS ion Digital Learning Hub is a digital
system these institutions excel in research and portal that collaborates with organizations to
development e.g. IIMs, NITs, IIITs. The offer courses across various fields. All the
Institutions inherently design for change in portals provide higher education courses
constantly shifting their processes to compete like graduate, post-graduate,
graduate, and research
under innovation conditions. courses.
The educational institution decentralizes Swayam a portal available free of cost online
decision making, develop procedure, broadly education portal runs on the principle of
define jobs and evaluate performance access, equity, and quality.
subjectively.
NPTEL portal is coordinated majorly by IIT Stable Product and Process Change: Mass
Madras along with other IITs has web and Production Strategy
video courses across 23 disciplines. The portal Most of the education institutions permit them
has recorded lectures from IITs, which are then to standardize courses, routines work, and
uploaded online for others to access. standard rules and procedure and centralize
Swayam Prabha is a group of 32 DTH decision making. The focus of these
channels, which run high-quality educational educational institutions is to strive for the
programs. The contents are provided by largest size and lowest cost structure in the
NPTEL, IITs, UGC, CEC, IGNOU, and education sector. These institutions often are
NCERT. The courses are available for students hierarchical and bureaucratic.
from class 9 to post-graduation level in the Classroom-based learning: Technology is an
subjects of arts, science, commerce, performing integral part of these institutions. The
arts, social sciences, agriculture among others. institutions made minor changes in student
Virtual Placements experience like animated content, usage of
PowerPoint instead of blackboard, and usage of
Virtual placements use virtual campus
gadgets like laptops, tablets to take down notes
recruiting tools to leverage the process, the best
[12].
way to start with is either live interviews,
placement talks, offer technical assessments, Dynamic Product Change and Stable
ensure that each stage of the process is Process Change: Mass Customization
managed virtually with real-time updates, and Institutions are facing customers making
to accurate the digital hiring process. unique and unpredictable demands. The basic
Virtual Internships approach of these institutions is to meet the
demands through flexible platforms; the focus
The Virtual internship programs allow
is on individual customer satisfaction through
participants, to gain work experience in a flexibility and responsiveness. The institution's
remote setting. The virtual internships allow strategy is to produce course variety rapidly
the intern to get first-hand experience with and inexpensively. The decision-making of
companies on a global scale while working these institutes is through centralized
from the comfort of their chosen location. The coordination.
student virtual but still part of a team, in Specialized degrees: The students are
constant communication with the team, and can interested in opting for industry-focused
manage/supervise the project. programs. So there is a demand to offer
specialized programs is picking up.
Online Learning Management System
International Collaborations: Indian
Many universities in India are integrating the Institutions and Students are looking for
online Learning Management System (LMS) collaboration with world institutes to excel in
platform and Enterprise Resource Planning collaborative research, and to showcase their
(ERP) into their web portal. Students can work and innovations.
remotely log in to access course material and
also attend live classes with teachers. Pre- Stable Product Change and Dynamic
recorded lectures, videos can be uploaded on Process Change: Continuous Improvement
the LMS & ERP platforms making it easy for The focus of the Institution is customer
students to go through them multiple times. satisfaction through process improvement.
LMS & ERP adoption is still poor in many These institutes manage rapid innovation and
parts of India where students do not have the use of the new process, systems, and structures
access to computers or broadband internet. to facilitate long-term organization learning
about products to make continuous
improvement efficient these institutions
employ cross-functional teams to improve the
process. The members of these teams turn to
that function-specific work and execute the and closely with teachers and colleagues from
rules they develop. across the world. Online learning is cost-
effective and even profit-making. However,
Invest in Online Learning
technical support for faculty, financial
Universities and colleges have started to invest management, professionalism, and a team
heavily in online teaching. Students want to approach to course development and delivery
grab the opportunity to work collaboratively are the factors for success.
video- Free Meeting Screen- Whiteboar Meeting E2E Plans
conferencing version participants sharing d recording encryptio from
apps available (default) n (p/m)
Zoom Yes 100 Yes Yes Yes No $14.99
Microsoft Teams Yes 250 Yes Yes Yes No $5.00
Google Meet Yes 100 Yes No Yes No $6.00
Cisco Webex Yes 200 Yes Yes Yes Yes $13.50
Platform-Based Learning Suite transformative teaching. Most of the

universities and colleges are into offering
A learning platform from a students'
virtual reality campus tours.
perspective, is a place to find educational
content, search and buy courses or whole Augmented reality (AR) drives students in
educational programs. And, from an educator’s enhancing their perception of real engagement
view, it is the place to deliver knowledge. by simulating superimposed, artificial objects
Online learning platform takes off a lot of the in real-world environments and. Students can
hard work of building an audience, marketing, explore, the experience they are present in that
and sales, on the control over pricing, user environment, or be involved in virtual objects.
data, and branding. Online learning platforms Today's students are accustomed to the
are often used to identify marketplaces for oscillation between an online and in-person
courses or education like Coursera, Udemy, experience. In the future, to improve student
and Lynda, while course platforms usually engagement and success rates teaching
refer to the software used to create a learning strategies of choosing the right technologies is
environment. going to matter.
AI and Big Data Universities be wise to Re-skilling /Up skilling through certification
embrace and lead AI in deep learning, which Technology changes and globalization are the
applies machine learning tools and techniques two possible reactions for individuals to re-
to solve any problem that requires “thought” – skill or up-skill. India is home to over 4,450
human or artificial. Harnessing machine Edtech startups. The likes of Toppr BYJU'S,
learning technology, much like the AI tools in Vedantu, and Unacademy are quite popular in
educational applications span from giving India. Global brands like Coursera, Udemy,
computers “vision,” to speech recognition, and LinkedIn Learning are also popular in the
machine translation, medical diagnosis, etc. Indian market.
The institutions need to believe that deep Quality Circle: Education is about quality.
learning is the new scientific infrastructure for The quality concept is applied to service
research and learning. improvements. The quality circle recommends
solutions. The curriculum is an important
Virtual Reality and Augmented Reality
aspect of quality education. Extension
Virtual Reality is a new reality tech that activities, Skilling, Placement, and Learning in
stimulates immersive experiences for colleges a broader context are of interest in forming
and universities. Virtual Reality can bridge the quality circles in universities and institutions.
gap between learners and educators. The future
of classroom experience will be hyper- Discussions
immersive blended with augmented reality, The recommendations for higher education
artificial intelligence, and new technologies in institutions during the pandemic
It is important for institutions to find out of education is allowing the big corporate to
solutions with the use of technology and conduct virtual placements and virtual
through integrated digital learning platforms, internships minimizing the impact of the
digital skills for teachers and students. pandemic on career advancement. Education
Ensure data privacy and security when sharing needs to be technology-driven in rural parts
that the platforms shall not violate the student of India through a policy with a proactive
data privacy. approach in enabling the digital landscape.
A blended approach to address psychological The learning tools, quality IT infrastructure,
challenges before teaching is necessary. faculty and student familiarization to
Creating online communities to ensure and teaching technologies strengthen the
examine the communication is happening countries digital learning. The institution
among the stakeholders. and a student in the interior part of the
Provide orientation in Training Teachers and country with the conduction of strategic
Students towards streaming of lectures. analysis can grow considerably well in the
Plan and Map the Curriculum with the distance market.
learning programs to meet the needs of
Limitations and Future Scope of Study
students. Formative tests, exercises in the
course can be used to assess the student. The Study is a theoretical study can be
validated and tested further using analytical
Conclusion
tools. The technology enabled education has to
A tech-embedded education system is a be assessed as the concept is new in India the
pandemic-proof solution that can also be studies can be carried in assessing the level of
something that can continue after the
student participation and his understanding
pandemic. It’ll allow the higher education
institutions and students to widen their while the higher educational instructions are
outreach, by s adapting to a digital aiming for transforming their approach from
curriculum and digital learning. The mode traditional to technology driven.
References
1. Dharmendra Nagda, and Dharmesh (2020) Influence of COVID-19

Motwani. Role Of Ict In Quality confinement on students’ performance in
Enhancement In Higher Educational higher education. PLoS ONE
Institutions- Opportunities & 15(10):e0239490.
Challenges,Vol 4, No 12 (2018) doi:10.1371/journal.pone.0239490https://j
International Education and Research ournals.plos.org/plosone/article?
Journal. id=10.1371/journal.pone.0239490
http://ierj.in/journal/index.php/ierj/article/v 4. Hadiya Habib “Role of ICT in Higher
iew/1797 Education” International Journal of
2. Du C, Zan MC, Cho MJ, Fenton JI, Hsiao Creative Research Thoughts (IJCRT) 2017
PY, Hsiao R, Keaver L, Lai CC, Lee H, IJCRT | Volume 5, Issue 4pp 2810-2813
Ludy MJ, Shen W. The Effects of Sleep 2017 | ISSN: 2320-2882
Quality and Resilience on Perceived https://ijcrt.org/papers/IJCRT1704371.pdf
Stress, Dietary Behaviors, and Alcohol 5. R. Krishnaveni and J.Meenakumari
Misuse: A Mediation-Moderation Analysis .“Usage of ICT for Information
of Higher Education Students from Asia, Administration in Higher education
Europe and North America during the Institutions – A study” International
COVID-19 Pandemic. Nutrients. 2021 Journal of Environmental Science and
Feb;13(2):442. Development, Vol. 1, No. 3, Pg 282-286,
DOI:10.37074/jalt.2021.4.1.10 August (2010), ISSN: 2010-0264.
3. Gonzalez T, de la Rubia MA, Hincz KP, https://www.semanticscholar.org/paper/Us
Comas-Lopez M, Subirats L, Fort S, et al. age-of-ICT-for-Information-
Administration-in-A-Krishnaveni- and opportunities from the perspective of

Meenakumari/c1a7381954044c0372a3504 students and instructors. Journal of
53e350f3f850628b9?p2df Computing in Higher Education. 2021
6. Rupal S. Patel, Renewing Use of ICT in May 3:1-8.
Indian Higher Education. GAP BODHI 10. Uttam Kr Pegu , Information and
TARU14 An International Peer-Reviewed Communication Technology in Higher
Open Access Journal of Humanities, Education in India: Challenges and
Volume: I, Issue: 1 ISSN: 2581-5857. Opportunities,. International Journal of
https://www.gapbodhitaru.org/res/articles/ Information and Computation Technology,
Paper%20%203%20-%20Copy.pdf ISSN 0974-2239 Volume 4, Number 5
7. Richard J. Schonberger. World Class (2014), pp. 513-518.
Manufacturing: The next decade: Building https://d1wqtxts1xzle7.cloudfront.net/5116
Power, Strength and Value, Free 2454/ijictv4n5spl_11.pdf?1483444494=&r
Press,2013, ISBN: 978-1476747347 esponse-content-
https://books.google.co.in/books?id=ojJNJ disposition=inline%3B+filename%3DInfo
oZ4_L0C&printsec=copyright&source=gb rmation_and_Communication_Technology
s_pub_info_r#v=onepage&q&f=false .pdf&Expires=1620998481&Signature=B
8. Tahir Hussain Bhat, Andleeb Nazir , and XcTnHyYTRGA2DZimUSa3xs2B7bG60
Afaq Alam Khan “Impact of ICT in 3OGYZ3cvl7T42RcNUulPzzLp~Mx-
Higher Education: Opportunities and pMTUSKqPrzl7at7EqHLg8R~eKI~jJSTF
Challenges” International Journal of CfNwb36UuS8o2zkfnl24Ad~2BvVG7yO
Advance Research in science and ubEIlP20SVBYDKWpTu8985yLbhv-
engineering” Vol.7, No.4,2008, Pg 1503- jQEjtC9ZEYwTJW71e8Y~5vuzZD07icr
1509. WexpRTa1Ozds5ysYbfQjOazQUAt~1xM
http://www.ijarse.com/images/fullpdf/152 8iMZWninoxbbcFveSmDfi86rS2-
4847045_ I21QNvDGH8BZM8-9M521aOiUn~x~oP4
JK1491IJARSE.p uHUrTeOCSW2RAj3uLRtuJ7-
9. Maatuk AM, Elberkawi EK, Aljawarneh bo9wy3Ghr4ID8V5pCg9YLzmuJw8FcH0
S, Rashaideh H, Alharbi H. The COVID- vi-Y8GgU9q56N9zjtg__&Key-Pair-
19 pandemic and E-learning: challenges Id=APKAJLOHF5GGSLRBV4ZA
Websites
1. https://www.orfonline.org/expert-speak/ to-Boost-Market-Growth-Technavio
challenges-quality-education-times- 6. https://sannams4.com/overview-of-
pandemic/ online-education-in-india/
2. https://www.expresscomputer.in/guest- 7. https://www.indiaeducationforum.org/rese
blogs/how-important-is-technology-for- arch-
education-in-india-during-the- report/Trends_and_Opportunities_in_the_
pandemic/61285/ Online_Learning_Market.pd
3. https://www.financialexpress.com/educati 8. https://www.techrepublic.com/article/zoo
on-2/coronavirus-impact-temporary- m-vs-microsoft-teams-google-meet-cisco-
school-closure-may-cost-india-over-400- webex-and-skype-choosing-the-right-
billion-says-world-bank/2104537/ video-conferencing-apps-for-you/
4. https://timesofindia.indiatimes.com/home/ 9. https://www.ibef.org/download/education
education/news/10-online-learning- -and-training-apr-2019.pdf
platforms-by-hrd-
ugc/articleshow/74964325.cms 10. https://www.researchgate.net/publication/
5. https://www.businesswire.com/news/hom 350849186_Concerns_of_College_Studen
e/20200710005233/en/Online-Education- ts_during_the_COVID-
Market-In-India-2020-2024-Increased- 19_Pandemic_Thematic_Perspectives_fro
Penetration-of-Internet-and-Smartphones- m_the_United_States_Asia_and_Europe
THE PUBLIC DISTRIBUTION SYSTEM (PDS) AND FOOD SECURITY IN INDIA:

SPECIAL REFERENCE TO COVID 19
A. Shamshad
Department of Political Science, AKI’S Poona College of Arts, Science and Commerce, Camp, Pune
ashamshad2002@gmail.com
_______________________________________________________________________________________
ABSTRACT
India’s Public Distribution System (PDS) is considered as one of the colossal distribution networks in the world. It came
into prominence as a rationing measure during World War II. Import of food grains was the important factor of
distribution through PDS before 1960s. Due to the scarcity of food, it accelerated its activities in the 1960s;Later, the
Food Corporation of India and Agriculture Prices Commission was instituted to better local accumulation and dearth of
food grains for Public Distribution System. PDS had bloomed into a more advanced global scheme for the disposal of
government funded food by the end of 1970s.This scheme was improved upon for the better access to the common masses
hailing from mountainous and remote regions particularly targeting less privileged section in the society. Keeping this in
account, government initiated the Targeted Public Distribution System i.e., TPDS in1997, with a focus on needy people.
The work of TPDS is centred on supplying the subsidised food grains such as wheat and rice via far-reaching network of
ration shops. Such procured food grains from farmers are delivered to the ration shops under TPDS for the proper
distribution of this entitlement to the beneficiaries. It’s the duty of centre and state to identify the poor, procure grains
and get it delivered to the beneficiaries. The National Food Safety Act (2013) was passed in the parliament in September
2013.This act functions on the existing TPDS to deliver the food to the needy households making it their justifiable and
legal right. Food Security and Public Distribution System (PDS) work independently. Food Security stands for
affordability, accessibility and availability of uninterrupted food supply to all the people. The problem of production and
distribution of food is detrimental to the poor households that are affected with food insecurity. In these types of cases,
the responsibility of food security is dependent on the proper scrutiny surveillance of government and the Public
Distribution System (PDS). As per the Rome Declaration on World Food Security at the World Food Summit which was
held in 1996, “Food security exists when all people have legal access to sufficient and nourishing food to fulfil their
dietary requirements and food preferences for healthy life”.
________________________________________________________________________________
Keywords: Public Distribution System, National Food Safety Act, Rationing, Food Security
Introduction to the economically backward classes of society

at affordable rates. It is binding on the Centre
British introduced Public Distribution of grains
and State Government to inspect and regulate
through “rationing” system in India during the
the Public Distribution System. Activities like
Second World War. Rationing system and
procurements, transportations, storage and large-
Public Distribution System (PDS) have played a
scale allocation of food grains fall under the
very pivotal role in eliminating the threats of
domain of Centre Government whereas, State
famines and provided massive food security to
Government is responsible for distributing these
the needy people. This system was initiated in
items to the disadvantaged section of society
Bombay in 1939, afterwards reached to other
through Fair Price Shops. Other activities like
parts in India. One of the aims of government in
identifying the families below poverty line
India was to provide the food security to all the
(BPL), allocation of food supply, issue of ration
citizens of their country since post
cards and monitoring of Indian food security
independence. Ministry of Consumer Affairs &
system i.e., FPSs are taken care by the State
Food and Civil supplies started Public
Government. These are established by the
Distribution System to fulfil the objectives laid
government of India under Ministry of
down in this regard. Whereas, the Targeted
Consumer Affairs, Food and Public Distribution
Public Distribution System was instituted in
in cooperation with State Governments. They
1997.
distribute subsidized foods and non food
Public Distribution System is a government
commodities to the disadvantaged people in the
sponsored body with far-reaching chain of
society in India. These commodities are
rationing shops which works towards the
comprised of rice, wheat, sugar etc by food
distribution of food and non food commodities
departments are distributed through certified
ration shops in several States across India. The period. Government helps the Public
Public Distribution System is acquired and Distribution System with the timely
maintained by a Government - owned supplies, incurs the cost of subsidy and
corporation i.e., Food Corporation of India. determines the quantity goods and rates to
Followings are the objectives and features of the be distributed per family.
Public Distribution System; 5. The primary objective of Public Distribution
System is to provide basic minimum share of
1. To safeguard the low-paid workers by
essential commodities at economical rates to
ensuring the supply of certain minimum
the most disadvantaged sections in the
quantities of food grains at low price or
society and accounts for the stabilization of
availing the welfare facilities (food at
their open market price to keep a tab on
affordable prices) for the poverty-stricken
sudden rise in the prices during crisis. The
people.
government charges low rates as compared
2. Ensuring equitable distribution.
to open market prices and lessens the
3. Controlling the inflation on essential
procurement and other costs borne by the
commodities in the open market for price
government.
stability.
6. Basically, it covers sensitive urban areas
4. Availability of rations during the dearth of
where scarcity of food grains and essential
food grains, and
commodities become huge political
5. Eagle-eyed observation on private trade.
obligations of Government.
With a view to relieve famine and draught
Public Distribution System is considered as a
victims Public Distribution System started its
significant food security network in exposure
work of supplying the food to them in 1939, it
and public expenditure. Rations shops cannot
has also accelerated its activities in larger areas
control the herculean task of food grain supplies
for the procurement, distribution, civic supplies
and meet the requirements of disadvantaged
and pricing policies of food grains. Followings
sections of society. The PDS has been criticised
are some of the important features of Public
for its unfairness in the equal distribution of
Distribution System:
commodities and is believed to be fiasco in
1. The work of Public distribution System is to serving the poorer strata of population.
distribute the selected essential commodities
via fair shops such as cooperative owned by Phases & Reforms in the Public Distribution
Government and rationing shops. These fair System in India
shops are controlled and directed under the Amid 1960s, management of scarce supplies
Government’s inspection. were emphasized in critical situation. As per the
2. Wheat, sugar and rice are distributed as funding avenue Stoppage of PL 480 (Public
major food grains. Other items like edible oil Law 480 also known as Food for Peace (and
and kerosene are also distributed to the low- commonly abbreviated PL 480) can be utilized
income group in the society. for foreign aid. The imports compelled the
3. The functioning of free market mechanism is Government to procure the grains internally. For
never hampered by the working of the the availing of food security, Indian took
Public Distribution System except in some significant initiative with regard to provide a
statutory rationing areas. Consequently, it more sustainable institutional framework. In this
becomes a dual economy. Customers had regard, establishment of FC and Agricultural
two options to purchase the food grains Prices Commission are termed as Agricultural
either from the open market or Fair Price Costs and Prices (BACP) Commission which
Shops. heralded the inception of this phase. As per the
4. Through the internal procurement system or BACP’s recommended prices, the FCI gains the
imports and buffer stock, the government food commodities to dispatch through PDS and
acquires the required amounts of food grains a share of the procured quantity is kept as
and other items by these two systems which ‘buffer stock’ to meet any unprecedented
are maintained with the demand of shortage situation. If farmers are benefited with
affordable prices, this will surely open up food situation. Therefore, there was a sharp
employment opportunities for poor farmers and increase in the net availability of food grains
allied workers, consequently, it will turn into a from 75 million tons in 1968 to 99 million tons
huge source of happiness for them. in 1977 and again it witnesses a sharp rise in
Major segments of this system were inclusive of later years reaching million tonnes in 1991.
unconventional set ups and methods of
Reforms in the PDS
procurement, storing and distributing of food
grains. There was a terrific progress in the food It has been noted that the Government has
security system as a pivotal part of brought several programmes to modernise the
developmental strategies to incorporate changes Public Distribution System. The government
in the selected food crops, specifically wheat made use of advanced technology such as
and rice. digitalisation and computerisation of Public
There was a surge in the food grains production Distribution System to expedite the project.
during 70s & 80s in the country. There was also Government wants to automate all the ration
a drastic increase in the accumulation of buffer shops across the country. This initiative has
stock. With this, the main emphasis on buffer already been started in the state of Gujarat and
stock maintenance and price control shifted to Odisha by their respective governments. Some
hike in Public Distribution System supplies. In states are making use of innovative technologies
the 4th plan, (1969-1974, Planning to ensure the distribution of food grains in
Commission) it states that “in so far as food transparent manner with the active involvement
grains are concerned the basic objective is to of panchayat and village leaders with proper
provide an effective Public Distribution System. monitoring from Gram Sabha.
The procured quantities were in excess as The Department of Food and Public Distribution
compared to the requirement of Public in collaboration with the Indian government
Distribution System needs and minimum reserve have sorted out critical facets of PDS that need
was maintained”. urgent improvement to perform its work
As compared to the requirement of PDS, the successfully. Followings are some of the
procured quantities of food grains were measures;
abundant and minimum reserve was maintained 1. Beneficiary identification, and addressing
a programme named Antyodaya were started to inclusion/exclusion errors
lessen the poverty and reduce the overstocking 2. Addressing diversions and leakages
of FCI godowns in fifth year- plan. 3. Managing food grain storage and ensuring
The imports gradually slowed down during this timely distribution
period in the year 1975, though small in 4. Effective accountability and monitoring, and
quantity, there was a net exports of food enabling community monitoring
commodities during this time. To maintain the 5. Mechanisms for grievance redressal
level of buffer stock, the imports were consistent 6. Ensuring food security
relatively with less quantity. Government
reinforced the PDS in this period for a stable Leveraging Aadhar
and enduring feature of strategy to regulate the One of the recent discoveries in our country is
prices, lessen the fluctuations in them and obtain Aadhar card which modernizes the Public
an equal distribution of consumer goods. By the Distribution System. Indian Government
end of 1970s, the PDS was mainly restrictive to considers the Unique Identification number i.e.,
city population and did not avail adequate food Aadhar a mandatory document for the exclusive
to the rural population in need. Some state verification of the records of its citizens across
governments extended the scope of Public the country. The official order for the UIDAI
Distribution System to rural areas and (Unique Identification Authority of India)
demarcated the target group during the late entails the definition of the usages of this unique
1970s and beginning of 1980s. These states are number for critical applications and services.
Gujarat, Andhra Pradesh, Tamil Nadu and UIDAI can play a vital role in the Public
Kerala. During the later years in 80s and 90s, it distribution System application. The
was also as there was well marked change in the
beneficiaries of PDS can be documented into the and food preferences for vital and healthy life.
Aadhar system. It is noted that amalgamation of Public Distribution System is the most important
Aadhar system with PDS will reduce duplicity medium through which the government ensures
and bogus beneficiaries in Public Distribution food security to the people at the micro level.
System database which will facilitate in But, the working of PDS and the other
reducing wastage and distractions in the system. government policies have not been successful in
Reinforcement of PDS not only needs achieving food security at the desired level in
digitalization of operations but also coordination India. Presently, huge modernization drive is
with Panchayat Raj Institutions and local needed to reach supply of food grains to the
community through transparent system. In this actual receipt without the leakages and diversion
system, a mechanized system will keep track the of the route to the grass level. There is also
flow of food grains and guarantee that there is urgency for the digitalization and
no human intervention in the handling of computerization of data with regard to
records. beneficiaries and entire food chain supply to
The principal goal of Government of India is to ensure the PDS works successfully. These
modernize the Targeted Public Distribution attempts would make the schemes crystal clear;
System (TPDS). TPDS plans to avail subsidized assist uproot fraudulent practices, leakages and
food through the large network of ration shops. allows the beneficiaries to get the products as
Food grains such as wheat and rice that are per their legal rights. It can be supped up after
availed under TPDS are obtained from farmers, the assessment of the Public distribution system,
apportioned to the states and delivered to the this process is subject to the numerous
ration shops to be taken by the appropriate challenges like diversion of food grains and
beneficiaries. It is an assiduous task as Public leakages, bogus ration cards, inclusion
Distribution System functions in all the States /exclusion errors, , lack of transparency and
and UTs than 5 lacks Fair Price Shops across practicality of Fair Price Shops.
varied operational environments. Antyodaya The PDS speeds up the supply of food grains
Anna Yojana (AAY) was started in December, and the dispatch of essential merchandises to
2000 for ten million poor families. It anticipates disadvantaged population through the large
providing food commodities at a subsidized rate network of Fair Shops regularly at the
of Rs.2/ kg for wheat and Rs. 3/ kg for rice. The subsidized rates throughout the nation. It is an
States/UTs are required to incur the distribution integral part of the Government’s policy to
cost, inclusive the margins to dealers and manage the food economy in the country. This
retailers as well as the transportation cost. Thus, scheme was started with an intention to regulate
the whole benefit of subsidy is being passed on the rising subsidized food bill incurred by the
to the legal beneficiaries under the Antyodaya government as well as to ensure a clearer
Anna Yojana (AAY). targeting of poor and the disadvantaged group.
Under the join responsibility of Central and
Challenges before Public Distribution System
State government, the Public Distribution
Despite the obsolete system of Public System works. Through the Food Cooperation
Distribution and tremendous stock of food of India (FCI), Central Government has
grains, starvation and malnutrition still engulf presumed its duty for storage, procurement,
various parts of India. These disparities across transportation and bulk allocation of food
states have deteriorated the matters further to the commodities to the State Government. These
sufferings of poor. At this very inception, India responsibilities are comprised of allocation
is confronting the problems of food insecurity within the State, issues of Ration cards,
for quite a long period of time. Food security is identification of suitable families and the
one of the fundamental criteria for the supervision and functioning of Fair Price Shops
improvement and formation of human capital (FPSs). Under the Public Distribution System
and the productiveness of human being. rice, wheat and kerosene are being apportioned
Precisely speaking, food security exists when all to the States /UTs for distribution.
people have physical and economic access, The major beneficiaries of PDS goods are rural
nutritious food to fulfil the dietary requirements families of six crore whom about 72 lakh tonnes
of food commodities were sanctioned during most of the functionaries under them in the
2006. The government introduced Targeted Department are typically callous and resort to
Public Distribution System (TPDS) from June, corrupt practices.” This remark has been given
1997. As per this scheme, the states are by the Central Vigilance Committee headed by
supposed to formulate and implement full- Justice D.P. Wadhwa, retired Supreme Court
fledged arrangements for the identification of judge, submitted its report to a Bench of Justice
poor. Taking into account the failure of PDS Dalveer Bhandari and K.S. Radhakrishnan,
with regard to the service of poor people, the hearing a petition on the need for streamlining
government began issuing unique ration cards to the PDS. Senior counsel Colin Gonzalves,
the people living Below Poverty Line. For appearing for the People’s Union for Civil
ensuring the food security to the poor, PDS has Liberties, announced the committee’s concerns
evolved one of the major instruments of the over the mal functioning of the PDS in various
government’s financial policy and is considered States, and submitted a chart based on hid
as the largest network of its kind across the findings. The Bench recorded the report for
globe with 5 lakh Fair Price Shops. further deliberation. The report stated that
However, the data points out that around 60% fraudulent practices in the PDs was a cancerous
poor households using PDS are unable to get growth and had to be abolished as the patchwork
ration supplies. It is also stated that in high would be in vain. “The Central government
poverty-stricken states, ‘out of stock’ scenarios gives a whopping Rs. 28,000 crores annually to
were as high as 80%. And 34% of those going to the States to subsidise food for the poor but till
PDS offices had to pay four or more visits the recommendations of the committee are put
before any action was taken against their in place, the poor will go on suffering at the
grievances. Moreover, around 50% of them took hands of corrupt officials, dishonest fair price
the help of bribe to get the new ration cards shop (FPS) owners, treacherous transporters
issued. Even though, the owners of fair price and, possibly to a large extent, unscrupulous
shops are to be blamed for the corruption, they millers as well.” There was diversion of food
also have genuine problems which also need to grain supplies to the PDS on a massive scale,
be addressed. Very meagre margin in business black marketing by Fair Price System dealers, a
accounts for low profitability, particularly after strong nexus between the government officials
the launch of Targeted Public Distribution and the dealers, improper records in registers,
System. They are given very low credit they and most importantly political interventions
cannot life up enough food grains from the hamper public distribution. Starting from Delhi,
Government godowns. Sometimes, they have to “the PDS is inefficient and corrupt. There is
pay the bribe to the PDS officials to get the diversion and black marketing of PDS food
sufficient supply of food grains. The grains on a large scale. Subsidised food grains
government-run Food Corporation of India do not reach the poor, who desperately need the
procures food grains for PDS from farmers. same. These poor people never get the PDS food
Many times, the support price is very low as a grains in proper quantity and quality. There is
result of that FCI can only bargain for low all-round complicity of FPS, transporters and
quality grains and the rest goes to the open corrupt officials of the Civil Supplies
market. Corruption is more rampant in UP, Department.”
Delhi, Bihar and the Northern State’s PDS, The report stated that in Rajasthan, Jharkhand &
whereas, Himanchal Pradesh and Kerala are Bihar the PDS “is unsatisfactory; there are many
considered as comparatively less corrupt states. irregularities, irregular lifting of grain, no lifting
Central Vigilance Committee appointed by the and bulk lifting. PDS has collapsed. The
Supreme Court has censured PDS as one of the committee was informed that political pressure
most corrupt sectors, mentioning political is put on officials and the advisory committee in
intervention as the root cause of its failure in appointment of FPS dealers.” The distribution
various states. “Corruption is all pervasive in the mechanism of these States has been continued in
entire chain involved in the PDS. It continues to the hands of the most corrupt and inefficient
remain a formidable problem. It is true that people. The committee was informed that rotten
some officers are doing a good job but then grain and good grain are lying together in the
godowns.” If FPS owners do not pay the globally due to this pandemic. The UN has
godowns manager Rs. 10-20 a bag, he supplies pointed out that we are going through one of the
rotten grain, says the report. The report also says worst phases in the history of impeding food
that in Bihar, “the most disappointing fact is that emergency. The World Food Programme has
food grains are not distributed every month. The estimated that around 130 million people will
committee visited many villages and towns and suffer from food insecurity. This number may
met lots of beneficiaries. It was a general rise to 820 million as predicted by the State of
complaint that during the last many years, they Food insecurity in the World report, 2019.The
got food grains only for 2 to 3 months. The state covid-19 pandemic and the Nation-Wise
of affairs in Karnataka, Orissa, Uttarakhand and lockdown has caused many people to lose their
Gujarat was similar, says the report. livelihoods and multiplied the demand
depression that the country was already
Global Hunger Index & Food Security
confronting. Due to the growing unemployment
As per the report, our country is ranked 94th rate and stagnant rural wages, the country was
among 107 countries in the Global Hunger already going through the worst phase long
Index (2020) and falls under serious hunger before the lockdown. Coronavirus pandemic
category with 27.2 score. Last year, it was 102nd deteriorated the precocity of many people’s lives
out of 117 countries. In the hunger index, India throughout the country. Consequently, the dual
is behind Nepal (73rd), Bangladesh (75th), attack of slowdown and rising unemployment
Pakistan (88th), and Indonesia (70) among other has resulted in massive food insecurity in India
countries. Among 107 countries, merely 13 during this time.
countries have done worse than India namely It’s a matter of great concern that this might
Rwanda (97th), Nigeria (98th), Liberia (102nd), have worst effects on the nutrition status of
Afghanistan (99), Chad (107), and Mozambique people. In this situation, PDS plays a vital role
(103rd) among others. Other 17 nations in extending support to the affected population
including, Turkey, Belarus, China, Ukraine, under the supervision of government. Even
Kuwait and Cuba shared top spots with GHI though these are some significant innovations to
scores less than five as per the website of the provide the food and cash to the people, still
Global Hunger Index that tracks the same. The there are some lacunas which need to be
mortality rate stood at 3.7 % under five years addressed. The availability of abundant food
old population. Report says that 14% of India’s stocks can be utilized as an opportunity to
population is affected with malnutrition. It also provide universal Public Distribution System
points out that the recorded a child stunning rate which is immensely needed now. Other issues
of 37.4 %. These children known as stunted such as inclusion of oils and pulses, universal
children are those who have “low height against food support through school meal, decentralized
their age”, signifying severe under-nutrition. procurement, Integrated Child development
Working of PDS in the period of Covid Services and enhanced cash transfer through
Pandemic social security schemes must be deal with
earnestness. Undoubtedly despite some
There is huge growing concern of food shortcomings, PDS has contributed a great deal
insecurity and economic slowdown due to the of support to the people after the worst impact
Coronavirus pandemic all over the world. of national lockdown. However, the necessity
Lockdowns, border closing, quarantine for some support in the form of subsidized or
requirements, social distancing restrictions have free food commodities to a substantial
hindered the supply and distribution agencies population than those documented by the
putting the huge pressure on them to make national Food Security Act (2013), is crucial in
availability and accessibility of food grains to current scenario where there is a huge spike in
the masses. Precarious occupations are worst hit Corona Virus cases which has devastated the
pushing the people to the loss of their economy.
livelihoods and even risking their basic needs. It evident from the data from last few months
As per the estimate of World Bank around 71 that PDS has been proven as one of the effectual
million people will confront utter poverty platforms which garnered support and access to
the people. Azim Premji University has The following points may be taken into
conducted a Covid-19 survey and have found consideration for the effective implementation
that about 20% of worst affected households did of TPDS:
not receive any rations, while about 50 % of 1. Commodities other than wheat and rice
households were not benefited from the cash needed to be kept out from the ambit of
transfer facility, consequently exhibiting the TRDS. Its major goal of providing
relative effectiveness of Public Distribution subsidized food to the poor is to be
System comparatively other relief measures. It guaranteed food security to them. Wheat and
has been traced in the research that in rice are two sought after commodities in
Chhattisgarh that more than 300 families were India which are considered as the basic
95% of them reported having ration cards, on necessities. Hence, food subsidy must be
the basis of these findings Public Distribution provided for these two commodities.
System must provide rations to the affected 2. Commodities such as sugar should be
families during this pandemic. excluded from the purview of PDS. Sugar
Concluding Observations & Suggestions should not be tempered with and the system
of tax on sugar should be stopped.
It has been found that there are large number of
3. It is expostulated that if the production of
beneficiaries in the single state that is, Kerala
coarse cereals is accelerated in dry land
that have a durable infrastructure for the PDS.
areas, it can check the degradation of soil
Given the state of Kerala, it has received around
leading to environmental damage. But there
12 percent of food grains through Public
is a difficulty in supplying coarse cereals
Distribution System. This state accounts for
through PDS on subsidized rate under the
around 3 percent of total population of India. In
cover of food security. Coarse grains cannot
this circumstance, it can be pointed out that
be stored for long time that makes it
Bihar is the less benefited states from the
unsuitable for the long-term storage and
operation of PDS, this state does not get any
distribution under PDS. Standard variety of
neither any benefit from food procurement and
coarse grains is not available; hence, it
distribution operations. India cannot afford to
cannot be included in the national level
rely on more import of food commodities to
program. But government can take some
meet its requirements due to its large continental
concrete steps to supply it as per the needs of
dimensions. On the other hand, there is an
the specific regions.
urgent need to produce more food grains and
4. One of the commodities is Kerosene oil
during the times of need, rest of the
which is supplied through PDS for the poor.
commodities can be imported for the efficient
This is one of the essential items in which a
distribution. The PDS in the country supports
large-scale illicit diversion occurs, poor and
the transfer of the food grains grown locally to
needy are cornered by fraudulent people and
the benefits of poor and disadvantaged
subsidized kerosene is adulterated with
population in several geographical regions. The
diesel to make money. Subsidy on kerosene
growing food subsidy and food stocks have
is meant for the poor but rich corner them
come under the suspicion for the cost-
and uses the same for commercial purposes.
ineffectiveness of the PDs. There is a need for
5. A large number of commodities should be
the reconstruction of the PDS to abolish hunger
resisted to be included in the coverage of
and make food commodities available to the
food subsidy.
poor in a cost-effective manner wherever they
6. The FPS (Fair Price Shop) should be
may be.
allowed to sell all the commodities other
Not only the major food deficit States but also
than wheat and rice at full market prices via
the major surplus States are the major
PDS outlets so as to ensure their economic
beneficiaries of PDS. Not only the PDS helps in
feasibility.
achieving the objectives of food security in food
7. The beneficiaries from the population below
deficit States but also it generates a ready
the poverty lines should be taken into
demand for the supplies in surplus producing
account for TDPS and food security. And
States.
the people a who are above the poverty line 10. There are various operational schemes which
should be facilitated with the availability of are in the favour of the welfare or income
good grains at a stable price in the market. transfer schemes where distribution of
The coverage of food subsidy should not be essential commodities is involved. Such
extended to the people above poverty line schemes with others which are serving the
ensuring no needy person is left from the
same purpose can be merged and some sort
benefits. Stability in food grain prices
should be ensured through the maintenance of convergences among them can be
of a buffer stock and open market operations strengthened.
of the FCI. Nevertheless, there is huge 11. Some community grain banks can by
surplus of food grains with FCI, it’s established in very sensitive areas where
advisable to continue the supply of food from disadvantaged people can take grain
grains under PDS to the APL population for when they are in need and pay off the same
the temporary relief. in the form of grains after their ordeal is
8. Given the liberalization of external sector, over. Government and the administration
timely exports and imports will supplement must work towards safeguarding them from
the operation of buffer stock leading to the
natural disaster like earth quake and flood
smaller size of required buffer stock.
9. For identification and other purposes rations which spurt starvation.
cards can be used by the administration. For At last, old, sick and disabled must be provided
this to work smoothly, multi-purpose with a minimal amount of social security as they
identity cards should be issued in future. are not able to give their services even if the
Many people produce ration cards only to
employment is available for them. Government
prove their identity before the
must implement some special schemes so that
administration.
they are not starved.
Reference
1. Angus Deaton & Jean Dreze,( 2009) Food Economic & Political Weekly, 48(45–46),
and nutrition in India: Facts and 55–60.
interpretations, Economic & Political 8. JeanDreze, Excess stocks of the Food
Weekly, 44(7), 42–65. Corporation of India must be released to the
2. Government of India, Agricultural Statistics poor, Indian Express,2020.
at a Glance, Ministry of Agriculture, New https://indianexpress.com/article/opinion/col
Delhi, 2003. umns/coronavirus-lockdown-food-for-poor-
3. Government of India, Annual Report 2002- migrants-mass-exodus-jean-dreze-6353790
03, Department of Food and Public 9. JeanDreze, Rohan Khera, Thought for food.
Distribution, New Delhi, 2003. Outlook, 2015. https://www.outlookindia.com/
4. Government of India, Report of the Working magazine/story/thought-for-food/295144
Group on Public Distribution System and 10. Kaushik Basu,( 2011) India’s foodgrain
Food Security, Planning Commission, New policy: An economic theory perspective,
Delhi,2001. Economic & Political Weekly, 46(5), 37–45.
5. Government of India, Tenth Five Year Plan 11. P.V. Rajeev, (1999) Planning for Social
(2002-07), Planning Commission, New Reforms, Deep and Deep Publications, New
Delhi, 2002. Delhi.
6. Jean Dreze & Rohan Khera, (2015) 12. Reforms in the Public Distribution System
Understanding leakages in the public
distribution system, Economic & Political and Better Targeting of Food Subsidies,
Weekly, 50(7), 39–42. Working Group on Reforms in the Public
7. Jean Dreze, Rohan Khera,( 2013) Rural Distribution System and Better Targeting of
poverty and the public distribution system. Food Subsidies during the 12th Plan Period,
Department of Food and Public Distribution, system: A national and international

Planning Commission, July 29, 2011. perspective, World Bank: Poverty and Social
13. Tata Economic Consultancy Services, Study Policy Department,1996.
to Identify Diversion of PDS Commodities 15. http://planningcommission.nic.in/aboutus/co
from Fair Price Shops, New Delhi, 1998. mmittee/wrkgrp12/pp/wg_pds.pdf.
14. World Bank, India’s public distribution
IMPACT ANALYSIS ON SHRIMP FARMS OF NAVSARI DISTRICT, GUJARAT, INDIA

WITH REFERENCE TO COVID 19 PANDEMIC
R. Bengani, K. Manoj, A. Varsani and J. Chotliya

Department of Aquatic Biology, Veer Narmad South Gujarat University Surat, Gujarat, India
Email: ranjanabengani@vnsgu.ac.in
_______________________________________________________________________________________
ABSTRACT
Shrimp farming sector has been known as an important area in the economic growth of India. On March 25th, India
initiated a strict 21-day lockdown, later extended till May 31st. Still at present India is not fully unlocked and
restrictions are present. The whole aquaculture production system was disturbed due to COVID 19 pandemic. Survey
was carried out using structured questionnaires through direct interview to collect the relevant information from
different shrimp farms of Navsari district. Survey results revealed that COVID-19 pandemic affected different areas of
shrimp farming sector. Overall farm production and development was affected including reduction in consumption,
cancellation and loss of export order, reduction in selling price as well as demand, delayed in stocking, lack of
technical support and skilled manpower. The most affected area was transport sector that hampered seed procurement
and selling of harvested crop. Inadequate storage facilities were principal reason for the spoilage of their production.
It is recommended to implement government assistance so farmers can sustain post lockdown for managing the farm
activities
________________________________________________________________________________
Keywords: COVID-19, aquaculture, analysis, production
Introduction households of coastal people. They are

involved in fry collection, farming, harvesting
Shrimp farming is an aquaculture
and processing of shrimp (Barmon, et al.,
entrepreneurial activity that exists in marine,
2006)
brackish or freshwater environments, culturing
India announced, a strict 21-day lockdown on
shrimp or prawn for human consumption.
March 25th and the lockdown was later
Shrimp are a highly nutritive valuable
extended until May 31st
commodity among the edible crustaceans
(https://fish.cgiar.org/news-and-updates,2020)
(Lemos et al., 2004). Globally, shrimps
All the fishing activities were halted and there
represent 15% of the world trade of fish and
was closure of storage facilities, markets, and
fishery products in terms of value, and are
processing plants. On April 9th, India’s
ranked as second highest food commodity after
Ministry of Home Affairs allowed fishing and
tuna (FAO, 2016).
aquaculture businesses to pursue their activities
In recent years, although global farmed shrimp
under conditions of following the COVID -19
production has increased to 5.7 million tons in
guidelines
2010 (Brito et al., 2016) and about 25% of
(economictimes.indiatimes.com,2020). The
shrimp sold on the world market are raised on
was great effect on both International and
shrimp farms (Boyd et al., 1998). Shrimp farms
national fishing sectors. Frozen shrimp which
operated with good management and business
makes up 70% of India’s seafood export
practices can be profitable and benefit the local
earnings have been severely affected
economy by creating jobs in production,
(https://fish.cgiar.org/news-and-updates,2020).
processing, marketing, feed manufacture,
or India’s shrimp farmers, the lockdown was
transportation and related support services. In
announced at the beginning of its summer
India farmed shrimps alone account for 51% of
farming season (between March and July). This
the total earnings and shrimp farming is being
period usually yields 60 percent of India’s
practiced in about 1.2 lakh ha with a
annual shrimp production, with the winter crop
production of 4.26 lakh ton (MPEDA, 2016). It
(between August and December) making up
has been recognized as a part of Blue
the
Revolution for geographic features of
remainder(https://thefishsite.com/articles/how-
southwest. It has created a wage – earning
employment opportunities to the poor peasant
has-indias-shrimp-sector-weathered
weathered-the-covid- 19 pandemic. Navsari district of Gujarat state
19-crisis) was selected purposively for the study as the
A majority of India’s farmers (85%) are small district accounts
unts for the largest shrimp farming
and marginal farmers with less than two area (12037.18 ha) in the state. There is total
hectares of land. More than nine million active 346 farms present in Navsari region. Farmers
fishers directly depend on fisheries for their of Navsari district mostly prefer to culture
livelihood, 80% of which are small scale Litopenaeus. Vannamei specie of shrimp which
fishers; the sector as a whole employs over 14 is of great economic importance.
million people
Materials
rials and Methods
(https://www.hakaimagazine.com/news/indias
(https://www.hakaimagazine.com/news/indias-
fishers,2020).The pandemic created an Basic information of farmers was collected
unprecedented economic, social and health from the Marine Products Exports
crisis with impacts on the most vulnerable Development Authority (MPEDA). Two
groups
oups including women (harvesters, blocks namely Jalalpore and Gandevi having
processors and vendors), migrant fishers, fish highest shrimp production in the district were
workers, ethnic minorities and crewmembers selected for the study (Fig.1). From each of the
(https://www.orfonline.org/expert
(https://www.orfonline.org/expert- sampled block, two clusters of villages were
speak/impact-covid19,2020). The study carried selected based on area under shrimp culture.
out by Bennett et al., 2020, on assessment on Total of fifty selected farms were visited and
COVID-19 19 impact indicated that shrimp details about farmers were collected. The
farming got disrupted as supply chains were farmers belonged to different villages like
broken, labour shortage emerged and market Machad, Matvad, KaradKaradi, Samapor, Onjal,
access got affected. Mangrol, Bhat, Bhinar, Danti, Mendhar, Vansi
The present study was aimed to identify
iden the and Borsi. Following the method of Kothari
constraints affecting aquaculture development and Garg (2014), personal interview of the
and profitability ratio of shrimp farming due to farmers was taken from the selected sites with
limitations and restrictions because of COVID-
COVID pre structured questionnaire.
Fig. 1 Map showing selected

sel blocks of Navsari district
graduation level (26%), Fig. 1.3. According to

survey it was observed that shrimp farmers
owned shrimp farming training to operate and
setup farm from different agencies. It was
found that 10% of farmer got their training
from Fisheries department & 84% from
MPEDA. (Figure 1.4). It was observed that all
shrimp farmers were experienced and around
60% of shrimp farmers had experience between
4 to 10 years followed by 14% farmers with
farming experience between 11 to 17 years, 12
% of the farmers had experience between 18 to
24 years and 14% of the farmers had
experience of 25 to 31 years (Figure 1.5).
It was observed that around 82% shrimp farms
were having farm size between 1 to 10 ha
followed by 8% with farm size between 11 to
20 ha. Around 6% of shrimp farms were
having a farm size of 21 to 30 ha and only 2%
of shrimp farms were having farm size 41 to 50
ha. (Figure 1.6).
Patil et al. (2018) in their study in Palghar
district, Maharashtra reported that higher
percentages of shrimp farmers (54.55%) were
with experience between six to ten years.
Similar results were reported by Kumaran et al.
(2017) mentioning that 94.65 % farmers were
with more than five-year experience in East
coast of India. Shrimp farming industry was at
its peak around 1994- 95 throughout India, but
collapsed thereafter and has established the
root once again after introduction of SPF
Litopenaeus vannamei in 2008. This may be
the reason that maximum shrimp farmers were
having experience between 6-10 years. The
present study findings are similar to that of
Dona et al., (2016) where they reported that the
education level of shrimp farmers indicated
that no one was illiterate. Similar results were
Fig 2. Map depicting selected villages also found by Srinivas and Venkatrayalu
(2016). Lekshmi et al., (2005) reported that 40
Result and Discussion percent had collegiate level of education in
shrimp farmers and 30 percent of shrimp
Basic Details of Farmers
farmers had primary education in Andhra
During the survey maximum numbers of Pradesh. Ogunmefun and Achike (2017) also
farmers were from village Samapor whereas reported that the majority (60%) of the
least number were from Machhad village respondents possessing secondary education
(Fig.1.1). The education level of farmers and 37% were found to possess tertiary
surveyed is summarized in (Table 1.2). education with only very few (3%) respondents
Education level of the respondents were indicating to have only completed primary
categorized into 3 group: secondary level of education.
education (38%). higher secondary (36%) and
Mohite (2007) reported that 65.79 % of the The observation regarding the overall
shrimp farmers had less than two hectares of production during the pandemic was accessed.
farm size. The results reported in the present It was revealed that only 10% of the farms got
study are in accordance with results reported good production whereas 64% of the farm
by Salunkhe (2018), Srinivas and Vankatraylu stated that the production was averages the
(2016), Randive (2008) and Gawade (2006). farms that got poor production due to lack of
Patil et al., (2018) in their study in Palghar infrastructure during pandemic was around
district, Maharashtra reported that most of 20% (Figure 1.10).
shrimp farmers (40%) were having farm area Due to COVID 19 pandemic India faced a long
between 2 and 5 ha. In South Konkan region, lockdown for 6 months. In these 6 months all
Maharashtra, most of the shrimp farmers were personal and government vehicle transportation
new entrants and immediate investment for services was disturbed. The same effect was
bigger size farm was not possible for them. also perceived in shrimp cultures as 10%
This could be the main reason for farmers revealed that inadequate transport facility was
opting small size farms. present and 78% of the farmers stated that there
Naik et al. (2020) reported most of the shrimp was high transportation cost. 56 % of the
farmers in South Konkan region, Maharashtra farmers said that there was inadequate
(52.54%) were registered with CAA, Similar manpower to provide transportation facilities.
findings were reported in the present study (Figure 1.11)
were The brief data collected on impact of COVID
19 pandemic revealed that uncontrolled
Covid 19 Impact Analysis in Shrimp Farms
disease occurred during the pandemic period
India went into a lockdown in late March 2020 that made them difficult to sustain in this field
to control the coronavirus pandemic. Many post pandemic. The questions on providing
farmers panicked and harvested their shrimp guideline to farmers to sustain during
while shrimp processing plants faced pandemic revealed that only 18% got some
cancellation and postponement of purchase type of guidelines and 16% of the farmers are
orders, possible shutdown of the factories and thinking of alternative livelihood (Figure 1.12).
migrant workers returning to their homes. In India aquaculture industry plays very
These all had challenges in receiving the important role in food chain and economy. The
harvested shrimp. lock down period because of COVID-19
The effect of COVID 19 was also perceived in pandemic brough great series of trouble with it,
shrimp farms of Navsari district. The impact of the aquaculture industry had to face various
COVID 19 was analyzed graphically. troubles relate to transportation and
The study indicated that 54% of the studied availability. According to Manoj et al. (2020)
farm concluded that the income was decreased after the pandemic lockdown, majority of
due to pandemic effect. 86% of the farmers farmers were having no active fishing days.
revealed that farm operation was affected while Farmers faced difficulties in different sector of
24% of the farmers claimed that that the overall farming process. Transportation problems were
farm development was affected due to faced due to high cost of transportation and the
pandemic, Figure 1.7. The effect of pandemic farmers were bearing huge losses due to lack of
on harvesting aspect was surveyed and it was proper storage facilities, transportation and
observed that COVID 19 had a great impact on prone to rot the production in the farm and
harvesting due to unavailability of accessories harbours. The absence of transport facilities
and workers. (Figure 1.8). The problem of limited the movement of migratory harvest
COVID 19 was also observed on post labour, technicians and supply of Agri
harvesting period of shrimp cultures. 48% of machinery. Stephens et al., 2020 also
the farmers revealed inadequate storage concluded that quarantine measures were the
facility, 74% stated inadequate ice facilities factors that severely affected the labour
and 34% of them stated that there was availability and other farm activities.
breakdown of cold chains (Figure 1.9). Complete lockdown in the harbours and the
landing centers has greatly affected the fisher-
folks’ day-to-day earnings in all coastal ,2021 in their survey reported adverse impact
districts (https://www.mssrf.org/content/ of COVID 19 pandemic on shrimp farming
covid19-impact-livelihoods). Kumaran et al sector.
Figure 1.1 Location of farm
Fig.1.5 Size of Farm
Figure 1.2 Education of Farmer
Fig.1.6 Number of ponds

Fig. 1.3 Training
Fig. 1.4 Experience of farmers Fig. 1.7 Overall impact of COVID-19 on farms
COVID-19
Fig 1.8 COVID-19 impact on accessories

Fig. 1.12 Impact on disease outbreak and other
factors
Conclusion
The present assessment on shrimp farmers of
Navsari district has clearly indicated that
COVID-19 lockdown and subsequent
disruptions in the supply chain movements
adversely impacted the activities in shrimp
aquaculture sector and led to a direct economic
loss to its different stakeholders. India's
Fig.1.9 Post harvesting impact due to COVID- dependence on the exotic SPF vannamei brood
19 stock from overseas suppliers will lead to a
long-term impact if the restriction on
international cargo movements continues for
longer period of time. Further, limitations
which forced the skilled technician and
farmworkers to be confined at home and in-
house migratory workers to leave for their
homes, undesirably affected all the components
of this sector and their daily source of revenue.
Farmers and hatchery operators are looking for
massive relief packages from the government
to minimize the negative economic impact and
Fig. 1.10 Impact on production due to COVID-19
sustain this dynamic agribusiness sector which
earned a lucrative profit last year. This
investigation has provided an initial assessment
of the pandemic based on shrimp farmers
perspectives. However, the situation is
dynamic and further assessments may be
required to fully understand the impact of this
pandemic on Indian shrimp aquaculture sector.
Fig. 1.11 Impact on transportation due to
References
1. Barmon, B.K., Kondo, T. and Osanami, F. K.K. Vijayan. 2021. Prospective impact of
(2006). Impact of Rice Prawn Gher Corona virus disease (COVID-19) related
Farming on Agricultural and Household lockdown on shrimp aquaculture sector in
Income in Bangladesh: A Case Study of India – a sectoral assessment,
Khulna District. Japanese Journal of Farm Aquaculture. 2021 Jan 30; 531:
Management, 43(1): 133-136. 735922. Published online 2020 Sep 11
2. Bennett N.J. Elena, Finkbeiner M., Ban 10. Lekshmi, P.S., Chandrakandan, K,
Natalie C., Belhabib Dyhia, Jupiter Stacy Kumaran, M., Balasubramani, N, (2005).
D., Kittinger John N., Mangubhai Socio economic profile of shrimp fimners
Sangeeta, Scholtens Joeri, Gill David, and its influence on the exdent of adoption
Christie Patrick. 2020. The COVID-19 of shrimp culture technologies. Fishery
Pandemic, Small-Scale Fisheries and technology 42 (2). 225-230.
Coastal Fishing Communities, Coastal 11. Lemos, D., Navarrete Del Toro, A.,
Management Cordova Murueta, J. H. F. (2004). Testing
3. Boyd, C. E. (1998). Water quality for pond feeds and feed ingredients for juvenile
aquaculture. Research and Development pink shrimp Farfantepenaeus paulensis: in
Series No. 43. International Centre for vitro determination of protein digestibility
Aquaculture and Aquatic Environments, and proteinase inhibition. Aquaculture
Alabama Agricultural Experimental 239: 307-321.
Station, Auburn University, Auburn, 12. Manoj K, Bengani. R and Varsani A.2020.
Alabama. Survey on Assessing the Impact of Covid-
4. Brito, C., Valle, B., Interaminense, J., 19 Pandemic on Aquaculture Sector, India,
Peixoto, S., Lima- Filho, J. V. and Soares, Journal of Interdisciplinary Cycle
R. (2016). Microbiological quality of Research,XII(IX),192-198
Litopenaeus vannamei culture using 13. Mohite Y T. 2007. Efficacy and
conventional and biofloc systems. constraints in Adoption of improved
Aquaculture research, 47(10): 3098-3108. aquaculture Practices by shrimp farmers in
5. Dona P, Sheela Immanuel, Ohja SN, Raigad District of Maharashtra. M.F.Sc.
Ananthan PS. (2016) Occupational Needs Thesis, Dr. Balasaheb Sawant Konkan
of Shrimp Farmers in Kerala. Indian Res. J Krishi Vidyapeeth, Dapoli, India. 58 p.
Ext. Edu September, 16(3). 14. MPEDA (2016). Species-wise fish
6. FAO (2016). The State of World Fisheries production, Marine Products Development
and Aquaculture. In: Contributing to food Authority. www.mpeda.org.in.
security and nutrition for all. Rome: 200 15. MSSRF. COVID-19 Impact on
7. Kothari, C. R., & Garg, G. (2014). Livelihoods of Marine Fishing
Research Methodology: Methods and Communities. Available online:
Techniques. New Delhi: New Age https://www.mssrf.org/content/covid19-
International Publishers. impact-livelihoods-marine-fishing-
8. Kumaran M, Ravisankar T, Anand P communities
R,Vimala D D and Balasubramanian C P. 16. Naik B., Patil, S.V. Shirdhankar. M,
(2017). Knowledge level of shrimp Yadav B, Tibile R , Chaudhari K,
Farmers on better management practices Wasave, S and Yewale, V.. (2020). Socio-
(BMPs) of Litopenaeus vannamei Economic Profile of Shrimp Farmers of
Farming: A comparative assessment of South Konkan Region, Maharashtra, India.
East and West coast of India. Indian International Journal of Current
Journal of Fisheries, 64(3): 93-99. Microbiology and Applied Sciences. 9.
9. Kumaran R. M. Geetha, Jose Antony, K.P. 1371-1380.
Kumaraguru Vasagam, P.R. Anand, T. 10.20546/ijcmas.2020.909.174.
Ravisankar, J. Raymond Jani Angel, 17. Ogunmefun SO, Achike AI. (2017)
Debasis De, M. Muralidhar, P.K. Patil, Socioeconomic Characteristics and
Constraints of Pond fish farmers in Lagos culture practices by the farmers of North
State, Nigeria. Agricultural Science Konkan region of Maharashtra
Research Journal; 7(10):304 -317. M.F.Sc.Thesis, Dr. Balasaheb Sawant
18. Patil S, Sharma A, Ojha SN, Shirdhankar Konkan Krishi Vidyapeeth, Dapoli, India.
MM and Dhaker HS. (2018). Emergence 206 -208p.
of shrimp farming and profile of shrimp 22. Stephens E.C., Martin G., van Wijk M.,
farmers in Palghar district, Timsina J., Snow V. (2020) Impacts of
Maharashtra.Contemporary Research in COVID-19 on agricultural and food
India, 9 (1):37-42. systems worldwide and on progress to the
19. Randive P C. (2008). Adoption of Shrimp sustainable development goals. Agric.
health management practices in the South Syst. ;183:102873.
Konkan Region of Maharashtra. M.F.Sc. 23. The fish.cgiar website.2020 [Online]
Thesis, Dr. Balasaheb Sawant Konkan Available: https://fish.cgiar.org/news-and-
Krishi Vidyapeeth, Dapoli, India. 116 p. updates/news/fish-and-aquatic-food-
20. Srinivas, D. and Venkatrayalu, Ch (2016) systems-covid-19-updates-india
Studies on present problems and prospects 24. https://www.hakaimagazine.com/news/ind
of shrimp farming in west Godavari ias-fishers-have-been-crushed-by-covid-19
district of Andhra Pradesh, India Advances 25. https://www.orfonline.org/expert-
in Applied Science Research 7(2) 49-54. speak/impact-covid19-rural-lives-
21. Salunke A B. (2018). Efficacy and livelihoods-india 64889/
constraints in adoption of Penaeus
(Litopenaeus vannamei (Boone, 1931)
EVOLUTION OF BIG DATA AND ITS FEATURE LEARNING USING BIG DATA
ANALYTICS
S.K. Swain
Centurion University of Technology and Management, Odisha, India
sangrambapun@gmail.com
_______________________________________________________________________________________
ABSTRACT
Now computing is totally based on generating and assimilation of data at an unusual pace. There are thousands of web
applications on the internet serving millions of masses round the world. These web applications are increasing their
content at an ever-increasing rate. We rely on the internet for any of our needs; be it searching for a meaning in
Google or shopping to communicate, which has resulted in such vast explosion of data. This over the explosion of data
is known as Big Data. Web applications not merely aim at providing services to the user, but also in a fast and efficient
manner with the increasing amount of data, it has become hard to handle, store and analyze the information. Big data
analysis has become very crucial part in the business point of view due to extraction of required information from huge
data collection.
________________________________________________________________________________
Keywords: Application Evolution, Feature Learning, Big Data Analytics.
1. Introduction across and within the establishments

collaboratively as well as in real-time.
The BDA history is intricately connected with
Big data consumes huge data volumes which
that of data science. The term “big data” is
are deliberated as unstructured, structured and
initially applied by Michael Cox and David
semi structured datasets that could not be
Ellsworth in 1997 in an IEEE conference
captured easily, stored, handled, examined,
paper to justify the data visualization and also
accomplished also accessible through
to justify the challenges presented by it on
customary software, hardware and database
computer systems. The rapid innovations and
management methodologies. The features of
improvement obtained in IT technology by the
big data are usually used to define it. (L.
end of the 1990s enables the origination of
Douglas) 3 proposed 3 measurements which
data in large volume. To highlight the
describes the opportunities including the
importance of analysis, collection,
challenges of enlarging big data volumes.
interpretation, and integration of business
Obviously describing big data and its features
information the business intelligence (BI)
possibly will be a continuing attempt, however
concept is created besides in what way this
still it will not have adverse impact on the
procedure set can aid businesses make certain
execution and handling of big data. Moreover,
determinations and acquire a clear knowledge
to describe big data the 3V’s have been used
about market activities and tendencies (Cox
continuously, the other dimensions of value in
and Ellsworth)1.
addition to veracity are included to explain the
The evolutionary phase for big data’s
data feature and integrity to suit 5 V's of big
development was from 2001 to 2008 (RE
Bryant) 2. At first the big data was stated data (P. Russom) 4. (Suthaharan S) 5 invented
based on its 3Vs (volume, velocity, and 3Cs: continuity, cardinality, and complexity
variety), then it evicted to be more complex arguing that the first 3 V's (volume, velocity,
application to fulfil the necessities to deal with and variety) will not favour the advance
the information explosion. Hadoop, Database detection of the big data features for
management systems and XML (Extensible classification.
Mark-up Language) Web services and are 2. Concepts and characteristics of Big Data
some of the Software and application
developments that adds analytical modules About the creation of Big Data several
and core function modules that concentrates descriptions are available which are based on
on augmenting the end users usability , and distinctive viewpoints, for example the
allows users to handle vast quantity of data perspectives which are the social movement
Special Issue on Recent Research Trends in Management, Science and Technology (August 2021 ) 305
perspective, the process-oriented, product Data. So as to highlight the processing issues

oriented and cognition-oriented. Presently, of Big Data, the mandatory technical structure,
Web technology advances and the increased particularly the scientific gadgets,
mobile and sensors devices bonded to the programming methods, statistical,
Internet have brought about the origination of computational and scientific innovations in this
massive data sets available on the Web which perspective is also emphasized.
are required to be stored and processed. Now Thus, (T Kraska) 9 described Big Data as: Once
the companies so as to support the decision for the usual solicitation of present methods
their strategies are starting realizing the doesn’t facilitate the consumers to gain
significance of using more data. The big data sensible, economical and valuable responses
for the companies in the globally integrated for data-driven queries. Likewise, (A Jacobs) 10
economy are considered as a business priority. mentioned Big Data as: Data that’s volume
A review on the Big Data has been presented makes us to view past sampled and exact
by (SK Swain et al.)6 that has determined to approaches which are dominant at that period.
form value about all faces in establishments, The veracity as introduced by (B Feldman et
workforce customers, or society as entire. al.) 11 is an alternative difficult feature. Few
Towards the NoSQL databases the new descriptions of Big Data are presently being on
paradigm is moved corresponding in a the basis of the behaviours of Big Data.
considerable way besides accessible computing Conversely, (SB Scruggs et al.)12 mentions that
platforms, open-source software, and the Big Data’s description is apart from the
commodity servers’ entrepreneurs, Researchers scope of these features besides must spread to
and businesses, likewise anticipated its ability “to be convenient and re-claimed,
innovations or vehemently point to concrete add value over time, as well as originate a
that may be related to the evasion group of multi-dimensional, systems-level
large data sets. At this scale the organizations understanding”. Data is developing at an
endure to collect more data, the big data exponential pace always on the internet. One of
analysis reinforcing process will become the difficult issues is to execute, save and
dominant. The business executives to remain examine this considerable volume of data in
competitive must adopt the innovative certain way that of least time-consuming. Big
techniques and developing technologies due to Data in industry and in skill is having more
big data. focus. The technology is shifting to data centric
The perspective which is product-oriented architecture and functional models. Basically,
highlights the qualities of data regarding with it’s necessary to determine the architectural
respect to their structures, speeds and sizes. elements, operative models, also the crucial
The inspiration of this perspectives is information models that formed Big Data
depending on a historic vision, to relate the Ecosystem. To explain all looks of the Big
number of data offered with the volume Data Ecosystem (SK Swain et al.)13 centres on
recently. National Science Foundation (N.N.I. the Architecture Framework of Big Data. This
Initiative) 7 refers the Big data as: Huge, has the modules such as Big Data
longitudinal, complex, diverse, or /and Infrastructure, Big Data Architecture
dispersed data sets created using sensors, tools. Framework, Big Data Definitions and
Today and in the future the available sources Analytics, Big Data structures, Big Data
are online transactions, click streams, emails, Lifecycle Management, and models.
videos and/many other digital sources. The perspective i.e. cognition-based, emphases
Moreover, the term big isn’t only define the on Big Data challenges regarding their
size. (MM Gobble) 8 expresses that data is big limitations and the cognitive capacities. The
due to its excessiveness, as of its fast range of commonly used software tools and
movement, or it is not organised suitably. The hardware environments is exceeded by the data
perspective which is process-oriented for capturing, managing, and handling it in an
emphasises the originality of processes endurable passed period to its user populace. A
involved and required in storing, handling, formative document from McKinsey Global
accumulating, analysing and searching Big Organisation highlights the same matter (J.
Manyika et al.)14: the size of data set is away feature of the data, the frequency of the data
from the capacity of classic database software creation and the requirement of creating real-
devices for capturing, analysing, storing and time outputs. The range and source of data
managing. Hence, Big data is conceptualized types are termed as Variety which emphasises
by this perspective through the statistic of several information sources and the distinctive
surpassing human capability for data schematics of all sources. E.g., the traffic
comprehending suitable outcomes to the dataset contains on roads vehicular traffics
mediation requirement for enhancing the numeric information and textual data about
ability of interpretation. At last, the perspective scheduled and active events (e.g., musical and
of social movement lures our attentiveness to sporting events) (Y.J. Xia et al.)19.
the break among the vision and realism, Moreover, some of the additional Vs, like
specifically the socioeconomic traditional and virtual, value, variability and veracity, have
ethical moves which underlie the Big Data’s been stated in literature for serving as
existence (H. Ekbia et al.)15. The vision of Big complement features of Big Data (H. Barwick)
20
Data has been emphasized by this perspective (C.W. Tsai) 21(P. Zikopoulos et al.)22: An
for technical innovation, ecological and organizations monetary worth is considered as
biomedical study, education and nation-wide ‘Value’ that can be obtain from Big Data
safety (OSP, Obama administration unveils) 16 processing. It contains two attributes: the large
and states Big Data as: The quantifying potential rate and very small density rate.
capacity that modify many features of Veracity relates to whatever range the data can
contemporary lift for transforming the art of be trusted because its sources reliability. For
management (A. Ignatius) 17, otherwise to be example, some devices may be more
portion of a main transforms which needs compromised when data’s are received from
general exertion. sensors. Instead defining by numerous Vs, (X
For conceptualizing the Big Data some of the Wu et al.)23 likewise emphasised the features of
perspectives present valuable insights and in Big Data features using the professed HACE
addition quite a lot of features can be reviewed. hypothesis: Big Data initiates with enormous -
The common measures volume, variety and volume, dissimilar, independent sources with
velocity are acknowledged and considered (D decentralized and distributed control, also
Laney et al.)18: data sets size is represented by aspire to deal with challenging and developing
Volume. The Big Data’s volume may be of relations between data.
terabytes or even petabytes, which is much
2.1.Big data definitions
apart the typical bounds of gigabytes and
megabytes. The in and out speed of data Table 2.1 shows the definitions of big data
indicates the velocity. It represents the dynamic which were given by NIST Standard.
Table 2.1 - Big Data Definitions
Volume Measures the available quantity of data to association, which certainly doesn’t
have to possess all of it only if it can retrieve it.
Velocity Measures the streaming, aggregation and data formation speed.
Variety Measures the intensity of representation data – text, images, video, audio, etc.
Veracity Measures the ability of data understanding – noise, biases, anomaly and so on.
Value Represents the data’s utility in decisions making.
2.2.The paradigms of Big Data science was prompted by this new paradigm.
A novel paradigm i.e., the fourth paradigm of The major problem of this data-intensive
science was shifted from calculation to data
scientific research was provided by the
and also from outcomes to thinking along with
behaviors of big data. The historical visions for
data. The curation, analysis, and capture were
the four existing scientific paradigms are
the three basic behaviors that were normally
shown in TABLE 2.3. A data- intensive
taken into consideration by (AJ Hey et al.)24.
The big data was processed to attain the cleaning were highly covered by the curation.
knowledge from data; this respective data was It also includes the policies and essential
obtained to help the intelligent decision. At the metadata to perform the integration and
same time, the data interpretation/visualization longevity. The data analysis process includes
was identified as an essential task. Based on some activities like analysis, databases
25
(Chen and Zhang) , the data utilization, and modelling throughout the
interpretation/visualization was considered as a workflow pipeline. A newly developing
separate task not like the analysis process. The database for a mentioned stream should give
existence of this data was obtained in all shapes solution to the 20 important queries that highly
and scales of resources (also include concerned by researchers is claimed by (AJ
individuals’ lives). A huge range of procedures Hay et al.)24.
like integration, representation, and data
Table 2.2 - Four Scientific Research Paradigms

Paradigm Time Methodology
Experimental Science 1000 years ago Natural phenomena depiction
Theoretical Science Last few 100 Newton’s laws, Maxwell’s expressions and so
years on.
Computational Science Last few decades Simulation of complex phenomena
Data-intensive Today Various sources are used to capture and
science generate data; data collaboration is supported
by Science, analysis, visualization, assessment
This table has been adopted from (AJ Hay et al.)24
When the big data was met by social science, environments and novel business practices.
a new discipline referred as computational The data spectrum obtained in this new
social science (or e-social science) was paradigm support the high value decisions by
developed. An enormous inter-discipline that developing a novel advanced technique, and
was involved in this computational social this data spectrum may extents the meso-level,
science analyses and apply data along with the macro-level to the micro-level.
aid of big data technologies and techniques 2.3.Big data applications
(R.M. Chang) 26. For instance, in the
computational organization science the The useful values were provided by this big
perspective of researcher’s were extended on data analysis procedure through suggestions,
organizational, social, and policy systems, judgments, decisions, or supports. An
then the computational models were enormous amount of applications were
leveraged, subsequently, it combines the included in this data analysis process, and
science, network, science, social, and these applications were found to be regularly
computer science. Likewise, in e-business, a changing and extremely complex. An
large number of novel methods were evolutionary process in data source is initially
developed for the research environments like reviewed in this section. Subsequently, the six
consumer behaviour, managerial and human major data analysis fields like text analysis,
decision-making process, market interactions network analysis, mobile analysis, structured
and operational processes. These methods data analysis, multimedia analysis and website
were developed by taking the valuable data analysis are then examined. Finally, several
gathering costs in the social networks key application fields of big data were
environment, digital entertainment and mobile introduced.
telephony, blogs, and also consider the 3. Application evolutions
innovative abilities which found difficult to
execute. By considering that, a paradigm shift In recent years, the big data analysis was
was developed in scientific research, due to introduced as an advanced analytical
the enforcement of organizational equipment, whereas it contains complex
programs and large-scale under particular text, and audio, etc. So, at that particular
analytical process. During the past decades, moment, a huge amount of advanced
the applications of data driven were technologies were applied for emerged data of
developed. For example, the BI was unstructured or semi-structured. For example,
developed as a prevailing technique as early as the valuable information from the images was
1990s, for various business solicitations and extracted by image analysis process, (e.g.,
also for network search engines in terms of face recognition). An automatic video
huge data mining process that was developed surveillance system for law enforcement,
in early 21st century. The influential and military, and business applications were
potential applications obtained from various applied with the multimedia analysis
sectors and its respective examination and technologies. Since 2004, various
data are also deliberated in below sections. opportunities like upload, create, and share
contacts were provided to each user by some
3.1.Evolution of Commercial Applications
online social media like online communities,
The business data obtained in earlier days was social multimedia websites, social networking
generally identified as a structured data. The services, Internet forums, and blogs.
companies gathered this data from legacy
system and store them in RDBMSs. An 3.3.Evolution of Scientific Applications
analytical techniques that are applied in this Scientific research that is performed in various
technique was developed in 1990s, whereas sectors like oceanology, environmental
this technique was found to be more simple research, astrophysics, and genomics attains
and intuitive e.g., in the forms of dashboard, huge data with maximum throughput
reports, search-basis business intelligence, instruments as well as sensors. The
data mining, interactive visualization, BIGDATA program was recently announced
predictive modelling, queries with condition, by the U.S. NSF (National Science
online transaction processing, and score cards Foundation) to elevate the efforts to extract
(RL Sallam et al.) 27. From the 21st century, a the insights and knowledge from huge
unique opportunity was delivered by the collections of complex digital data. Platforms
WWW (World Wide Web) and networks for of Big data were developed and valuable
organizations to have direct interactions with results were obtained by few scientific
the customers and also to have an online research disciplines. For example, in biology
presentation. The website mining and text physical computing resources, virtual machine
analysis process perform some process like resources, data service, network infrastructure,
customer trade analysis, market structure inter- operative analysis software,
analysis, product layout optimization, and coordination environment, and inter-operative
product suggestions. analysis software were applied by iPlant (SA
Goff et al.)28 to assist the students,
3.2.Evolution of Network Applications
researchers, and educators, in inspiring plant
The WWW and email services were highly sciences. Huge varieties like experimental
provided by the former generation of the data, reference or specification data,
Internet. Some of the process like web page observation data, model or analog data, and
analysis, Text analysis, and data mining were other derived data were included in this iPlant.
provided to perform the mining process on e- Over the previous couple of years, the big data
mail contends and also to develop a search has attained huge importance in various
engine. The applications that are developed in industries. It cannot be stored and process by
recent days are highly web-based, irrespective former databases as it includes large and huge
of their arena as well as design goals. The data sets. Big data process has the high
significant proportion of the global data capacity to handle such huge and large data
volume was accounted by the network data. sets because the large dataset was processed
Web was identified as a mutual stand for during the processing phase in a required time.
interrelated pages, full of different types of Text analysis was still identified as infancy,
data like videos, interactive contents, images, but it was found to be a promising one.
Because in various companies, 80% of data performance than the other classic approaches.
were obtained in unstructured form, where The researchers of this article telling that the
various analysis processes were work only for performance of this method is weak only
the structured data. when using the smaller hash codes and
additionally it is not suitable for sparse
4. Feature learning on big data
features, otherwise this method is better and
Here, some of the works based on the feature executes good for deep features. They planned
learning respect to big data which are related to make their method to support sparse
to our work is illustrated. features and improve the performance on their
(Min Chen et al.)29 presented a method for future work.
learning the features of medical image data. A new scheme of feature selection that will be
The researches of this article introduced a optimal for the learning of big data is
structure respect to deep learning for feature presented by (Adrian Barbu et al) 31. The
learning of medical image with unlabelled method used in this research was put on for
data. This framework efficiently learns improving different loss function,
features of medical images with the minimal arrangement, ranking and application in
amount of labelled data. The method proposed regression. The researchers telling that their
in this article is used for some task like method is very simple and easy for
classification, similarity check, and lung implement. The over fitting problem solved by
nodule recognition. From the experimental enforcing the second order prior on piecewise
results of this method shown that their method linear retort function. The experimental
is much fine than that of other approaches. assessment of this paper was performed with
The executions of auto-encoder approach were some existing methods and they yield better
good, but it was affected by the quality of performance on regression, classification and
data. This issue is caused by ignoring the role ranking and their method is highly scalable
of expert. On their future work, they planned and efficient.
to overcome this issue by integrating the For the feature learning corresponding to big
domain knowledge and data-driven feature. data in industrial IOT (Internet of Things)
The task like image retrieval, image (Qingchen Zhang et al.)32 presented a model
classification and object detection and of adaptive dropout deep computation with
segmentation are handled on big data with crowdsourcing. The main advantages of the
various methods. (Jamil Ahmad et al.)30 model in this article was, preventing the over
presented an article for classifying the features fitting of hidden layers, the parameter training
of multimedia big data. Here the authors and technique of crowdsourcing. These able to
extract the features and transform it into hash build an efficient distribution function and
codes by using the sensitivity property. This also it will monitor the available training
hash codes uses neighbour search procedures samples. The results of this model exhibited
for the retrieval of multimedia data from big that the offered model got finer execution than
data. Two steps were used in this approach. former crowdsourcing algorithms. The
Initially on first step, the introduced feature framework introduced in this paper
selection method chooses the features which successfully avoided the over fitting and
they have maximum diversity. They found efficiently collecting the samples of parameter
more than 1800 features from the 4000 training for feature learning in IoT big data.
features of images on their experiment. On the On their future work, some best methods are
second step of the proposed system, with the planned to apply on their model for improving
selected feature the fast Fourier transform is the performance.
estimated and higher frequencies will be (Dacheng Tao et al.)33 presented an outline for
binarizes with mean frequency. The method of improvement of image superiority in big data
this article aims to describe the selected images. The method recommended for this
multimedia feature as a signal and the feature article, takes five factors of image into
vector is estimated with the FFT. The account for the enhancement process. The
performance of this method earned higher authors of this article extract more than 15
features from the image and they estimated the computation to study the features on big data
visual quality of image with the regression in this the tensor was applied to make the
module. These are analysed and learned from complex connections of heterogeneous data.
big data sets. These big data sets were huge in Here, the tensor destination was applied in the
size than the similar image data sets. The layer of output by way of the average sum-of-
researchers of this article conducted squares error term of the renovation error to
experimental setup of their model with nine completely learn the basic data distribution. A
datasets and performance are measured in HBP (high-order back-propagation) algorithm
three categories. The performance of their was suggested in this method for training the
method is much better than FR and RR parameters. The existing BP algorithm was
algorithms. But the IQA of the proposed extended from the vector space to the high-
method is slightly lower than FR and RR order tensor space for developing this HBP
algorithms. On their feature work, they algorithm. Here, four various datasets were
improve the IQA frame for increasing the applied to perform the comparison process
performance of their method. along with existing multimodal deep learning
An intelligent fault diagnosis process was and stacking auto-encoder models. The
identified as a most promising tool in experimental outcome of this feature learning
mechanical big data, as it has high ability to methods was found to be more efficient when
provide accurate results for diagnosis process evaluated with SANE, STL-10, INEX, and
and also contains the capacity to process the CUAVE datasets.
gathered signals rapidly and effectively. The The deep learning merits were applied for
existing diagnosis methods, manually extract feature learning by (YBengio et al.)36 whereas
the features based on diagnostic expertise and this deep learning was obtained as the layered
prior knowledge. An important merit of this architecture as same as the human brain. The
traditional method was human ingenuity, but simple topographies were extracted by deep
these methods were found to be labor- learning with the help of raw data, likewise
intensive and time-consuming. An artificial via multiple layers; the more complex features
intelligence approach applied by the concept were learned (K. Hwang et al.)37. By
of unsupervised feature learning to study the performing the continuous optimization
topographies from raw data was highly process in both backward and forward
inspired by (Y Lei et al.) 34, so a two-phase propagation, a considerable number of
learning method was introduced in this features were obtained at the final stage by
method to perform the intelligent diagnosis multi-iteration learning. This feature learning
process in machines. In the initial phase, process was categorized into two types they
sparse filtering and an unsupervised two- layer are unsupervised as well as supervised
neural network was implemented to study the learning. The illustration data was transmitted
features directly from the signals which are from an input for prediction process to the
obtained from mechanical vibration. layer of top via supervised learning. The
According to this learned features, the health connection parameters that were obtained
condition was classified in the second phase between several pair of layers were optimized
by applying the softmax regression. The by applying BP. This optimization process
performance evaluation of this process was was performed by reducing the cost function
performed by applying a locomotive bearing value among the predicted and target value.
and motor bearing dataset respectively. This The most familiar neural network-based CNN
diagnosis method with motor bearing dataset (Y. Le Cun et al.)38 were applied to represent
shows high accuracy and superior the features through supervised learning.
performance in fault diagnosis than the other CNN was frequently implemented in speech
traditional approaches. Involvements of recognition (M. Chen et al.)39, text analysis,
learning technique reduce the human labor and image analysis (M. Chen et al.)39, etc.
and perform the fault diagnosis process more Particularly, in the image analysis field, CNN
easily while handling the big data. provide huge success, in face recognition (Y.
(Q Zhang et al.)35 introduced a model of deep Sun et al.)40, scene parsing (C. t) 41, cell
segmentation (H. Su) 42, neural circuit one by one, and also the significance of
segmentation (D. Ciresan et al.)43, analysis of several features was estimated individually
images the breast (P. Fonseca) 44 (A. R. (He et al.)63, (Zhang et al.) 64, but feature
Jamieson) 45 and brain lesion segmentation (K. correlation removal was identified as a major
Kamnitsas et al.)46 (A. Patel et al.) 47. For drawback. The classical method performs the
instance, a novel 3D-CNN was introduced to feature selection by concurrently manipulating
classify the poly candidates on circulating both feature correlation and discriminative
tumour cell (CTC). In (T. Uemura et al.)48 (W. information process. Maximum discriminative
Shen et al.)49 (X. Li et al.)50, and advanced features were selected by UDFS
convolutional networks were introduced to (Unsupervised Discriminative Feature
categorize the lung nodules from medical Selection) (Yang et al.) 65 for representing the
images by performing supervised feature data; furthermore, it also deliberated the
learning process. (M Gao et al.) 51 and (T manifold structure. The discriminative
Schlegl et al.) 52, some of the CNN-based information were normally encoded with
feature learning methods to categorize the labels for feature selection, so, the prediction
lung tissues based on the CT images obtained of good cluster indicators as pseudo labels was
from lungs. found to be essential to perform the
The unlabelled data present in unsupervised unsupervised feature selection.
learning process was applied for feature A large number of studies were performed on
learning process, whereas only a minimum the intelligent fault diagnosis process of
amount of data were applied to fine-tuning the rotating machinery intending to promptly deal
parameters, they are as RBM (restricted with the immense fault data and the accurate
boltzmann machine) (G. E. Hinton) 53, deep diagnosis outcomes were provided
belief network (G. E. Hinton et al.)54, auto automatically. Frequently, ANN (artificial
encoders (Y. Bengio et al.)55 and stacked auto neural networks) based approaches were
encoders (P. Vincent et al.)56. An auto encoder applied along with these studies, which for
process for pulmonary nodule classification features extraction use signal processing
and unsupervised feature learning were approaches and for classifying faults, the
proposed by (D Kumar et al.)57. To analyse features are given as input to ANNs. However
the breast images, a convolutional these approaches successfully functioned in an
autoencoder process was presented by intelligent fault diagnosis process of rotating
(Kalleberg et al.)58, and similarly, a RBM- machinery, but there are two defects exist. (1)
based design was developed by (Li et al.)59 for The features are extracted manually according
lung tissue classification. The analysis process to the sufficient previous understanding
in lung CT was proposed by (Tulder et al.) 60 regarding signal processing methods and
along with convolutional restricted Boltzmann diagnostic capability. Furthermore, concurring
machines. to a specific diagnosis issue these manual
A large number of algorithms for features are extracted and is undoubtedly inapt
unsupervised feature selection was introduced for other concerns, (2) The shallow based
to chosen the informative features from ANN frameworks were included in this
unlabelled data. The feature selection process process, as it restricts the ANNs ability of
was identified as a commonly applied learning the difficult non-linear interactions in
approach in unsupervised feature learning. fault diagnosis concerns.
The features that are conserving the data In artificial intelligence as an innovation, deep
likeness or having manifold structure designed learning maintains the ability to deal with the
from the entire feature space was selected above- mentioned shortages. Through deep
(Zhao and Liu)61, (Cai et al.)62, however, it learning, deep NNs (DNNs) employed the
fails to indirectly include the discriminative deep architectures, in preference to shallow
information within the data, so it cannot be ones, might be determined to collect the
applied directly in this shapelet learning valuable information from the valued complex
problem. Previously existing unsupervised non-linear functions and the raw data and
feature learning algorithms adopt the features estimated complex non-linear functions. An
innovative intelligent technique was delivered considering the proposed frameworks first
by (F Jia et al.)66 based on DNNs to handle the part, numerous methods for tackling the data
deficits of the aforesaid intellectual diagnosis inequity difficult, and then collective radial
approaches. The datasets obtained from basis kernel SVM and for solving the
planetary gearboxes and rolling element drawback of big data imbalance the SMOTE
bearings was applied by this method to (Synthetic Minority Over-sampling
authenticate its effectiveness. These datasets Technique) classifier is employed. Also, with
include immense measured signals concerning conventional data imbalance solver methods
distinctive health statuses under several such as under-sampling and SMOTE the
operating conditions. From the outcomes it’s proposed approaches oversampling
realized that the proposed technique even performance is compared. In this frameworks
though adaptively mines accessible fault second phase, the classifier output of SVM is
physiognomies from the obtained signals, replaced instead of target value of dataset to
likewise attains excellent diagnosis accuracy become stable in nature. In the final phase, the
related with the prevailing approaches. logistic regression were trained to perform the
Inefficient and unreliable human analysis is automatic pattern recognition process, the
replaced by the intelligent fault diagnosis modified dataset is used moreover for
techniques, intensifying the effectiveness of predicting the faults by means of a steel plate
fault diagnosis. The capability of multilayer manufacturing dataset in the RHadoop
nonlinear mapping highly supports the deep programming situation.
learning process to improve its accuracy in an To detect the fault in real-time and to identify
intelligent fault diagnosis process. A novel the industrial plants, a two-stage algorithm has
method named as Deep Convolutional NNs been offered by (B Costa et al.)69. The work is
(DCNNs) was suggested by (W Zhang et al.)67 fully dependant on the assessment of specific
along with WDCNN (Wide First-layer characteristics utilizing a novel evolving
Kernels). As input this approach utilizes raw classifier procedure as well as recursive
vibration signals (to generate more inputs data density evaluation. The suggested detection
augmentation is used), and the wide kernels at phase is dependent on the data space density
the initial convolutional layers were applied formation, and it isn’t as good as probability
for smothering high frequency noise and also density function, however is a suitable
for feature extraction. Small convolutional measure to detect anomaly/outliers. A Cauchy
kernels in the previous layers were used to function expresses this density and can be
support multiple layer nonlinear mapping. The determined in a recursive manner, which
AdaBN model was updated to enhance the improve its computational as well as memory
competence of domain modification. power capable and, so, relevant to on-line
For automatic pattern recognition, (A Kumar solicitations. A self-developing classifier
et al.) 68 offered a fault diagnosis according to system or auto class is functioned based on
agenda by enduring the challenges that were fuzzy rules and it is proposed in the detection
presented by data inequity problem obtained or diagnosis phase. When first rule base exists,
in big datasets by employing RHadoop Auto Class can develop or enhance it on the
programming atmosphere. The Fault diagnosis basis of the newly arrived faulty state data.
performed in the production method is a real- With the functioning time improvement of the
world example where the class-imbalance sulphur hexafluoride (SF6) electrical devices,
topic is extremely suitable. The regular the distinct amount of release possibly will
operating behaviour was presented by happen inside the device. Due to this the
majority of information that obtained from effectiveness of equipment’s insulation will
system whereas, faulty operating performance weaken and produce severe impairment on the
is restricted. The upholding approaches were device. So, this one is of viable impact for
condition-based method don’t function analysing fault and assessing state of SF6
accurately on such datasets and therefore, it is electrical device. In the current ages, data
challenging to form consistent patterns for the acquisition monitoring regularity of SF6
precise fault diagnosis. To deal with this issue, electrical devices has been always enhanced
and the gathering space has always been 5. Conclusion

extended, thus enormous data can be collected
Because of the numbers of methods are
in the substation database.
contributing to big data, day to day there is an
A two-level model for fault diagnosis was
introduced by (Hongxia et al.)70 to instantly increase in the popularity of big data is
handle enormous SF6 electrical device happened. For facing the various necessities of
condition monitoring data. At initial, the users, providing a perfect classification
monitoring data is pre-processed in the method for big data is essential. In many fields
previous training of the fault diagnosis model, numerous issues are solved by using various
for different missing values the first and approaches of big data. As already explained
different data filling methods are in the literature survey more number of
implemented. Then, on the Hadoop platform researches are performed in various fields of
the fault diagnosis algorithms are parallelized.
big data. Now the present research is given
At last, the time taken for the fault diagnosis
according to the research gaps. Sequentially,
of SF6 electrical equipment in stand-alone
mode and cluster mode is related by the purposes of the research have been
simulation, and the benefits of cluster mode in attained. The present research and its
processing massive data are verified. outcomes will provide future direction to other
researchers.
6. References
1. Cox, Michael, and David Ellsworth. Science and Mathematics (IJESM) 8, no 6

"Application-controlled demand paging (2017).
for out-of-core visualization." In 7. N.N.I. Initiative, Core techniques and
Proceedings. Visualization'97 (Cat. No. technologies for advancing big data
97CB36155), pp. 235-244. IEEE, 1997. science and engineering (BIGDATA)
2. Bryant, R.E., Katz, R.H., Lazowska, E.D., (2015).
2008. Big-data computing: creating 8. M.M. Gobble, Big Data: the next big thing
revolutionary breakthroughs in commerce, in innovation, Res. Technol. Manag.
science, and society computing. (2013): 64–66.
Computing Research Initiatives for the 9. T. Kraska, Finding the needle in the big
21st Century. Computing Research data systems haystack, IEEE Intern.
Association (Available at Comput. (2013): 84–86.
http://www.cra.org/ccc/files/docs/init/Big_ 10. A. Jacobs, The pathologies of big data,
Data.pdf). Commun. ACM 52 (2009): 36–44.
3. L. Douglas."3D Data Management: 11. B. Feldman, E.M. Martin, T. Skotnes, Big
Controlling Data Volume, Velocity and data in healthcare–hype and hope, Dr.
Variety" (PDF). Gartner. Retrieved 6 Bonnie 360 degree, Bus. Dev. Digit. Heal.
February 20. 2012 (2013) 122–125
4. P. Russom. “Big Data Analytics.” TDWI http://www.riss.kr/link?id=A99883549.
Best Practices Report, Fourth Quarter 12. S.B. Scruggs, Karol Watson, Andrew I.
2011. TDWI Research. 2011. Su, Henning Hermjakob, J.R. Yates 3rd,
5. Suthaharan, S. (2014). Big Data M.L. Lindsey, P. Ping, Harnessing the
Classification: Problems and Challenges in heart of big data, Circ. Res. 116 (2015):
Network Intrusion Prediction with 1115–1119.
Machine Learning. Performance 13. Manas Ranjan, Senapati Sangram Keshari
Evaluation Review, 41(4): 70-73. Swain, Srinivas Prasad, A Review Paper
6. Manas Ranjan Senapati, Sangram Keshari On Architecture Of Big Data , Journal of
Swain, Srinivas Prasad, A Review of Big Advanced Research in Big Data
Data and Its Current Research Directions, Management System (IJARBMS) 1, no
International Journal of Engineering, (2017): 1-10.J. Manyika, M. Chui, B.
Brown, J. Bughin, R. Dobbs, C. Roxburgh, 63 (2014): 67–80

A.H. Byers, Big Data: The Next Frontier 26. Sallam RL, Richardson J, Hagerty J,
For Innovation, Competition, and Hostmann B, Magic quadrant for business
Productivity. Report, McKinsey Global intelligence platforms. CT, Gartner Group,
Institute, 2012. Stamford (2011).
14. H. Ekbia, M. Mattioli, I. Kouper, G. 27. Goff SA, Vaughn M, McKay S, Lyons E,
Arave, A. Ghazinejad, T. Bowman, V.R. Stapleton AE, Gessler D, Matasci N,
Suri, A. Tsou, S. Weingart, C.R. Wang L, Hanlon M, Lenards A et al, The
Sugimoto, Big data, bigger dilemmas: a iplant collaborative: cyberinfrastructure
critical review, J. Assoc. Inf. Sci. Technol. for plant biology. Front Plant Sci 34(2)
66 (2015): 1523–1545. (2011): 1–16.
15. OSP, Obama administration unveils “big 28. Chen, Min, Xiaobo Shi, Yin Zhang, Di
data” initiative: Announces $200 million Wu, and Mohsen Guizani. "Deep features
in new R&D investments (accessed learning for medical image analysis with
December 2015). convolutional autoencoder neural
16. A. Ignatius, from the editor: big data for network." IEEE Transactions on Big Data
skeptics, Harv. Bus. Rev. 10 (2012): 12- (2017).
12. 29. Ahmad, Jamil, Khan Muhammad, Jaime
17. D. Laney, 3D Data Management: Lloret, and Sung Wook Baik. "Efficient
Controlling Data Volume, Velocity and Conversion of Deep Features to Compact
Variety, Research Note 6, META Group, Binary Codes using Fourier
2001. Decomposition for Multimedia Big Data."
18. Y.J. Xia, J.L. Chen, C.H. Wang, IEEE Transactions on Industrial
Formalizing computational intensity of big Informatics (2018).
traffic data understanding and analysis for 30. Barbu, Adrian, Yiyuan She, Liangjing
parallel computing, Neurocom- puting 169 Ding, and Gary Gramajo. "Feature
(2015): 158–168. selection with annealing for computer
19. H. Barwick, The “four Vs” of Big Data. vision and big data learning." IEEE
Implementing Information Infrastructure transactions on pattern analysis and
Symposium (accessed December 2015). machine intelligence 39, no. 2 (2017):
20. C.W. Tsai, ‘‘Big data analytics: A 272-286.
survey,’’ J. Big Data, vol. 2, no. 1, pp. 1– 31. Zhang, Qingchen, Laurence Tianruo Yang,
32, 2015 Zhikui Chen, Peng Li, and Fanyu Bu. "An
21. P. Zikopoulos, C. Eaton, Understanding Adaptive Droupout Deep Computation
Big Data: Analytics for Enterprise Class Model for Industrial IoT Big Data
Hadoop and Streaming Data, McGraw- Learning with Crowdsourcing to Cloud
Hill Osborne Media, 2011. Computing." IEEE Transactions on
22. X. Wu, X. Zhu, G.-Q. Wu, W. Ding, Data Industrial Informatics (2018).
mining with big data, IEEE Trans. Knowl. 32. Gu, Ke, Dacheng Tao, Jun-Fei Qiao, and
Data Eng. 26 (2014): 97–107. Weisi Lin. "Learning a no-reference
23. A.J. Hey, S. Tansley, K.M. Tolle, The quality assessment model of enhanced
Fourth Paradigm: Data-Intensive Scientific images with big data." IEEE transactions
Discovery, Microsoft Research, Redmond, on neural networks and learning systems
WA, 2009. 29, no. 4 (2018): 1301-1313.
24. C.P. Chen, C.-Y. Zhang, Data-intensive 33. Lei, Yaguo, Feng Jia, Jing Lin, Saibo
applications, challenges, techniques and Xing, and Steven X. Ding. "An intelligent
technologies: a survey on Big Data, Inf. fault diagnosis method using unsupervised
Sci. 275 (2014): 314–347. feature learning towards mechanical big
25. R.M. Chang, R.J. Kauffman, Y. Kwon, data." IEEE Transactions on Industrial
Understanding the paradigm shift to Electronics 63, no. 5 (2016): 3137-3147.
computational social science in the 34. Zhang, Qingchen, Laurence T. Yang, and
presence of big data, Decis. Support Syst. Zhikui Chen. "Deep computation model
for unsupervised feature learning on big 44. A. R. Jamieson, K. Drukker, and M. L.

data." IEEE Transactions on Services Giger, “Breast image feature learning with
Computing 9, no. 1 (2015): 161-171. adaptive deconvolutional networks,” in
35. Y. Bengio, A. Courville, and P. Vincent, SPIE Medical Imaging. International
“Representation learning: A review and Society for Optics and Photonics, vol.
new perspectives,” IEEE transactions on 8315, pp. 601–613, 2012.
pattern analysis and machine intelligence, 45. K. Kamnitsas, C. Ledig, V. Newcombe, S.
vol. 35, no. 8, pp. 1798–1828, 2013. Joanna, D. Andrew, K. David, R. Daniel,
36. K. Hwang, M. Chen, “Big Data Analytics and G. Ben, “Efficient multi-scale 3D
for Cloud/IoT and Cognitive Computing,” CNN with fully connected CRF for
Wiley, U.K., ISBN: 9781119247029, accurate brain lesion segmentation,” in
2017. Medical Image Analysis, vol. 37, pp. 61-
37. Y. LeCun, L. Bottou, Y. Bengio, and P. 78, 2017.
Haffner, “Gradient-based learning applied 46. A. Patel, S. van de Leemput, M. Prokop,
to document recognition,” Proceedings of B. Ginneken, and R. Manniesing,
the IEEE, vol. 86, no. 11, pp. 2278–2324, “Automatic Cerebrospinal Fluid
1998. Segmentation in Non-Contrast CT Images
38. M. Chen, P. Zhou, G. Fortino, “Emotion Using a 3D Convolutional Network,”
Communication System,” IEEE Access, SPIE Medical Imaging. International
IEEE Access, DOI: Society for Opticsand Photonics, vol.
10.1109/ACCESS.2016.2641480, 2016. 10134, pp. 201–206, 2017.
39. Y. Sun, X. Wang, and X. Tang, “Deep 47. T. Uemura, H. Lu, H. Kim, R. Tachibana,
learning face representation from T. Hironaka, J. Nappi, and H. Yoshida,
predicting 10,000 classes,” in Proceedings “Classification of Polyp Candidates on
of the IEEE Conference on Computer CTC Based on 3D-CNN,” in International
Vision and Pattern Recognition, pp. 1891– Forum on Medical Imaging in Asia
1898, 2014. (IFMIA 2017), pp.103-105, 2017.
40. C. Farabet, C. Couprie, L. Najman, and Y. 48. W. Shen, M. Zhou, F. Yang, D. Yu, D.
LeCun, “Learning hierarchical features for Dong, C. Yang, Y. Zang, and J. Tian,
scene labeling,” IEEE transactions “Multi-crop convolutional neural networks
onpattern analysis and machine for lung nodule malignancy suspiciousness
intelligence, vol. 35, no. 8, pp. 1915–1929, classification,” Pattern Recognition, vol.
2013. 61, pp. 663-673, 2016.
41. H. Su, Z. Yin, S. Huh, T. Kanade, and J. 49. X. Li, Y. Kao, W. Shen, X. Li, and G.
Zhu, “Interactive cell segmentation based Xie,“ Lung Nodule Malignancy Prediction
on active and semi- supervised learning,” Using Multi-task Convolutional Neural
IEEE transactions on medical imaging, Network,” SPIE Medical Imaging.
vol. 35, no. 3, pp. 762– 777, 2016. International Society for Optics and
42. D. Ciresan, A. Giusti, L. M. Gambardella, Photonics, vol. 10134, pp. 241–247, 2017.
and J. Schmidhuber, “Deep neural 50. M. Gao, U. Bagci, L. Lu, A. Wu, M. Buty,
networks segment neuronal membranes in H.-C. Shin, H. Roth, G. Z. Papadakis, A.
electron microscopy images,” in Advances Depeursinge, and
in neural information processing systems, 51. R. M. Summers et al., “Holistic
pp. 2843–2851, 2012. classification of ct attenuation patterns for
43. P. Fonseca, J. Mendoza, J. Wainer, J. interstitial lung diseases via deep
Ferrer, J. Pinto, J. Guerrero, and B. convolutional neural networks,” Computer
Castaneda, “Automatic breast density Methods in Biomechanics and Biomedical
classification using a convolutional neural Engineering: Imaging & Visualization, pp.
network architecture search procedure,” in 1–6, 2016.
SPIE Medical Imaging. International 52. T. Schlegl, S. M. Waldstein, W.-D. Vogl,
Society for Optics and Photonics, vol. U. Schmidt-Erfurth, and G. Langs,
9414, pp. 2801–2808, 2015. “Predicting semantic descriptions from
medical images with convolutional neural analysis with convolutional restricted

networks,” in International Conference on boltzmann machines,” IEEE transactions
Information Processing in Medical on medical imaging, vol. 35, no. 5, pp.
Imaging. Springer, pp. 437–448, 2017. 1262–1272, 2016.
53. G. E. Hinton, “Training products of experts 61. Zhao and Liu, 2007] Zheng Zhao and
by minimizing contrastive divergence,” Huan Liu. Spectral feature selection for
Neural computation, vol. 14, no. 8, pp. supervised and unsupervised learning. In
1771–1800, 2002. ICML, pp. 1151–1157, 2007.
54. G. E. Hinton and R. R. Salakhutdinov, 62. Cai et al., 2010, Deng Cai, Chiyuan Zhang,
“Reducing the dimensionality of data with and Xiaofei He. Unsupervised feature
neural networks,” Science, vol. 313, no. selection for multi- cluster data. In KDD,
5786, pp. 504–507, 2006. pages 333–342, 2010
55. Y. Bengio, P. Lamblin, D. Popovici, H. 63. He et al., 2005] Xiaofei He, Deng Cai, and
Larochelle et al., “Greedy layer-wise Partha Niyogi. Laplacian score for feature
training of deep networks,” Advances selection. In NIPS, pages 507– 514, 2005.
inneural information processing systems, 64. Zhang et al., 2015] Peng Zhang, Chuan
vol. 19, p. 153-160, 2007. Zhou, Peng Wang, Byron J Gao, Xingquan
56. P. Vincent, H. Larochelle, Y. Bengio, and Zhu, and Li Guo. E- tree: An efficient
P.-A. Manzagol, “Extracting and indexing structure for ensemble models on
composing robust features with denoising data streams. IEEE Transactions on
autoencoders,” in ACM Proceedings of the Knowledge and Data Engineering,
25th International conference on Machine 27(2):461–474, 2015.
learning, pp. 1096–1103, 2008. 65. Yang et al., 2011] Yi Yang, Heng Tao
57. D. Kumar, A. Wong, and D. A. Clausi, Shen, Zhigang Ma, Zi Huang, and
“Lung nodule classification using deep Xiaofang Zhou. l2, 1-norm regularized
features in ct images,” in 12th. IEEE discriminative feature selection for
Conference on Computer and Robot unsupervised learning. In IJCAI, pages
Vision (CRV), pp. 133– 138, 2015. 1589–1594, 2011.
58. M. Kallenberg, K. Petersen, M. Nielsen, 66. Jia, Feng, Yaguo Lei, Jing Lin, Xin Zhou,
A. Y. Ng, P. Diao, C. Igel, C. M. Vachon, and Na Lu. "Deep neural networks: A
K. Holland, R. R. Winkel, N. Karssemeijer promising tool for fault characteristic
et al., “Unsupervised deep learning applied mining and intelligent diagnosis of
to breast density segmentation and rotating machinery with massive
mammographic risk scoring,” IEEE data." Mechanical Systems and Signal
transactions on medical imaging, vol. 35, Processing 72 (2016): 303-315.
no. 5, pp. 1322–1331, 2016. 67. Zhang, Wei, Gaoliang Peng, Chuanhao Li,
59. Q. Li, W. Cai, and D. D. Feng, “Lung Yuanhang Chen, and Zhujun Zhang. "A
image patch classification with automatic new deep learning model for fault
feature learning,” in 2013 35th Annual diagnosis with good anti-noise and domain
International Conference of the IEEE adaptation ability on raw vibration
Engineering in Medicine and Biology signals." Sensors 17, no. 2 (2017): 425.
Society (EMBC), pp. 6079–6082, 2013. 68. Kumar, Ajay, Ravi Shankar, Alok
60. G. Tulder, M. Bruijne M,“ Combining Choudhary, and Lakshman S. Thakur. "A
generative and discriminative big data MapReduce framework for fault
representation learning for lung CT diagnosis in cloud-based manufacturing."
69. Costa, Bruno Sielly Jales, Plamen (2015): 289-303.
Parvanov Angelov, and Luiz Affonso 70. Miao, Hongxia, Heng Zhang, Minghua
Guedes. "Fully unsupervised fault detection Chen, Bensheng Qi, and Jiyong Li. "Two-
and identification based on recursive Level Fault Diagnosis of SF6 Electrical
density estimation and self-evolving cloud- Equipment Based on Big Data Analysis."
based classifier." Neurocomputing 150 Big Data and Cognitive 3, no. 1 (2019): 4.
EMPIRICAL STUDY OF RELATIONSHIP OF WORKING CAPITAL MANAGEMENT

AND PROFITABILTY OF INDIAN AUTOMOBILE COMMERCIAL VEHICLES
MANUFACTURING INDUSTRY
A. Raval1, A. Dave2, A. Joshi3 and T. Dave4

1,2,3,4
School of Liberal Studies, Pandit Deendayal Energy University, Raisan, Gandhinagar, Gujarat, India
1
avaniraval04@gmail.com, 2ashvin_dave25@yahoo.co.in, 3Ashish.joshi@rediffmail.com,
4
tndaveahmedabad@gmail.com
_______________________________________________________________________________________
ABSTRACT
This research study aims to examine working capital management practice carried out by six commercial vehicles of
automobile industry listed on BSE/NSE. The study tries to undertake the data evaluation of selected variables for the
tenure of ten years i.e. 2010-2019. The aim of this study is get to know association between dependent and independent
component. Dependent Variable is a return on equity and Independent Variables is a Fixed Assets Turnover Ratio,
Long Term Debt Equity Ratio and Current Ratio. Matrix of Co-efficient of Correlations, Descriptive and Regression
has been used for analysis. It is found that there is positively significant association between Fixed Assets Turnover
Ratio and return on equity. The result shows that rest of the variable has negatively significant relationship between
long term debt equity ratio and return on equity and rest of the two variables has positively insignificant relation
among Fixed Assets Turnover Ratio, current assets and current ratio.
________________________________________________________________________________
Keywords: commercial vehicles, profitability and working capital management.
Introduction and normal. The financial manager should be

alert of such requirements and finance them
Financial manager must see and pass his time
speedily. Regular growth in sales needs extra
to every day internal operation, relating to
investment in fixed assets. The finance
current assets and current liabilities of the
manager should pay specific courtesy to the
companies. Financial managers have to take
levels of current assets and the financing of
care to get the maximum profit because its
current assets.
huge portion of their valuable time is dedicated
to working capital problems. One of the most Literature Review
important things represents by investment in Princy N.K, Sreekrishnan P. (2017) analysed
current assets i.e. the total investment in asset. the performance of ULCCS Ltd. for the time
Financial manager should pay special care to period of five years with the span of 2011-12 to
the management of current assets on a regular 2015-16. The variables considered Current
base. Working capital management faces very Ratio, Quick Ratio, Absolute Liquid Ratio, Net
critical condition for all companies particularly Working Capital Ratio, Cash To Current
for small companies. Normally small Assets Ratio, Cash To Working Capital Ratio,
companies invest in current and may not much Inventory To Networking Capital Ratio,
invest in fixed assets. In India small companies Current Assets To proprietors Fund Ratio,
faces serious problem of collecting their debts. Current Assets To Fixed Assets Ratio,
Working capital management in the role paly Working Capital Turnover Ratio, Fixed Assets
of current liabilities in financing current assets Turnover Ratio, Current Assets Turnover
is far more important in case of small Ratio, Total Assets Turnover Ratio, Inventory
companies. As Different for large companies, To Current Assets Ratio, Gross Profit Ratio,
they face problems in increasing long term Net Profit Ratio, Cash To Sales Ratio,
finances.There is direct association among a Inventory To Sales Ratio, Inventory To
company’s growth and its requirement of Turnover, Inventory Holding Period, Gross
working capital management. Requirement of Profit Margin Ratio, Gross Working Capital To
the company’s for sales growth to invest more Sales Ratio. They are using ratio analysis trend
in debtors and inventory. The sales grow analysis, schedule of changes in working
regularly then requirement become very fast capital and common size balance sheet. Result
indicates that the efficiency of working capital Research Methodology

management is worthy. This study has measured only six commercial
Ben Le (2019) examined 497 companies for vehicles of automobile industries over a time
the period of one decade with the span of 2007 period of ten years i.e. 2009 to 2019. The
to 2016 has been taken for analysis. researcher evidence was acquired from Capitaline
consider some component i.e., Age, capital database. The aim of this study is get to know
expenditure, cash dividend, cash conversion association between dependent and
cycle, days payable outstanding, days sales independent component. Dependent Variable is
inventory, days sales outs tanding, government a Return on equity and Independent Variables
ownership, growth, interest, leverage, market is a Fixed Assets Turnover Ratio, Long Term
to book ratio, net working capital, return on Debt Equity Ratio and Current Ratio. Matrix of
invested capital, sales, size and stock return Co-efficients of Correlations, descriptive and
volatility etc. It was concluded that a regression has been used for analysis.
significantly negative association among Net
working capital and companies valuation, risk Objective
and profitability. The objective of this study is to know and
Dr. Aniruddha das and Debasree dutta (2019) assess the extent of effect and the relationship
examined three firms in automobile industry among dependent and independent variable of
listed on NSE firms over a period of five years the business enterprise in the commercial
with the span of 2012-13 to 2017-18 has been vehicle automobile industry.
taken for analysis. They had taken current
ratio, quick ratio, inventory turnover ratio, Hypothesis
debtor’s turnover ratio and asset turnover ratio H0: Fixed Assets Turnover Ratio has no effect
as variables. They were using ratio analysis for on Return on Equity
analyzing the data. It was found that Maruti H1: Fixed Assets Turnover Ratio has effect on
Suzuki India ltd. had constantly increased their Return on Equity
debtor’s turnover ratio while rest of the two H0: Long Term Debt Equity Ratio has no
firms debtors turnover ratio had decrease. effect on Return on Equity
Mahdi Salehi, Nadia Mahdavi, Saeed Zarif H1: Long Term Debt Equity Ratio has effect
Agahi Dari, Hossein Tarighi (2019) Examined on Return on Equity
91 companies listed on the Tehran stock H0: Current Ratio has no effect on Return on
exchange over a time period of eight years with Equity
the span of 2009 to 2016 has been taken for H1: Current Ratio has effect on Return on
analysis. The surplus stock return as dependent Equity
variable; finance and investment in working
capital and returns ratio to share value, Interpretation
dividends paid and changes in cash as 1. Table – A obtain the regression coefficient.
independent variables; and earning before It is specifying that direction; standardized
deducting invest and taxes, financial leverage Beta, significance level, values and VIF are
and financial cost as control variables. The remark in Table – A. Table – A displays
applied descriptive statistics, correlation that Fixed Assets Turnover Ratio, Long
matrix, multi collinearity analysis and Term Debt Equity Ratio and current ratio
conclusive statistics for analysis. The study as independent variables has association
found that there was a negative & significant with Return on equity. The β of Fixed
association among in cash and stock excess Assets Turnover Ratio as presented in
return. It was notice that there was no Table – A, stands at +0.256 viz. indicate
relationship among changes in working capital that Fixed Assets Turnover Ratio has
and stock surplus return. positive relationship with Return on equity.
The value of significance level is 0.001. It
is notice that the significance coefficient
beta (β) is statistically significance.
Therefore null hypothesis (FAR) is rejected hypothesis (CR) be accepted and H1 be

and H1 be accepted. rejected.
2. Table – A measured the cost of Long Term 4. Table – B indicate that the matrix of co-
Debt Equity Ratio i.e. -0.826 it is revel that efficient of correlations between
β has negative relationship with Return on independent variables i.e. FAR, LTDER
equity. The value of significance level is and CR have been considered by karl-
0.000 it means β is statistically clearly person coefficient of correlation. It is
significant. Thus null hypothesis (LTDER) precisely displays that all component of co-
be accepted and H1 be rejected. efficient of correlations’ value has not more
3. The β of Current Ratio as presented in than +0.4.
Table – A, stands at -0.162 viz. point out 5. Table – C obtainable the element of
that Current Ratio has negative relationship descriptive statistics. The outcomes
with Return on equity. The value of presented this studied will be more suitable
significance level is 0.036 It means that the to the organisations whose data sets
significance coefficient beta (β) is resemble the descriptive method indicate in
statistically significance. Therefore null Table – C.
Table – A: Regression Co-efficients
Regression t Significance
Co-efficients Level
Direction Value
Constant 0.214
FAR + 0.256 3.529 0.001
LTDER - 0.826 -11.245 0.000
CR - 0.162 -2.148 0.036
Dependent variable: ROE

Independent variables: FAR, LTDER, CR
Table – B: Matrix of Co-efficients of Correlations

FAR LTDER CR
FAR 1.000 -0.421 0.467
LTDER -0.421 1.000 -0.487
CR 0.467 -0.487 1.000
Table – C: Descriptive statistics

Variables ROE FAR LTDER CR
Mean 0.126 3.37 0.37 1.11
Std. Deviation 0.30 2.84 0.53 0.47
Findings 3. The Current Ratio has positive association

with profitability and significance
1. It is finding that there is insignificantly
coefficient beta is statistically relevant.
positive association among Fixed Assets
Turnover Ratio and Return on equity and
Conclusion
the significance coefficient beta (β) is
statistically significant. The research disclosed relationship between
2. Further find that Long Term Debt Equity Fixed Assets Turnover Ratio, Long Term Debt
Ratio has negatively association with Equity Ratio and current ratio and return on
Return on equity and significant co- equity that there is positively significant
efficient beta is statistically significant. association between Fixed Assets Turnover
Ratio and return on equity. The result shows
that rest of the variable has negatively two variables has positively insignificant
significant relationship between long term debt relation among Fixed Assets Turnover Ratio,
equity ratio and return on equity and rest of the current assets and current ratio.
References
1. Dr. Aniruddha Das, D. D. (2019). Working 3. Le, B. (2019). Working capital

Capital Management and Cash Flow management and firm’s valuation,
Analysis of Selected Indian Companies in profitability and risk Evidence from a
Automobile Industry. International Journal developing market. International Journal of
of Management, Technology And Managerial Finance, 15(2), 191-204.
Engineering, 9(3), 1144-1151. 4. Mahdi Salehi, N. M. (2019). Association
2. Princy N. K, P. (2017). A Study on between the availability of financial
Working Capital Management of Uralungal resources and working capital with stock
Labour Contract Co-Operative Society surplus returns in Iran. International
(ULCCS) Ltd. IJEDR, 5(3), 1179-1182. Journalof Emerging Markets, 14(2), 343-361.
BLOCKCHAIN AND DIGITALIZATION OF LAND ADMINISTRATION USING

DISTRIBUTED LEDGER SYSTEM
S. Charith G1, H. Reddy S2 and R.K. Rathode3

1, 2, 3
Department of Computer Sciences Engineering, Bharat Institute of Engineering and
Technology, Hyderabad
1
saicharith2005@gmail.com
_______________________________________________________________________________________
ABSTRACT
Blockchain is the one among the most important technical invention in the last two decades. This technique has paved
way to many tech giants to combat the problem of security and digital identity issues. Blockchain technology includes
cryptography and digital voting has a vast number of applications. Blockchain itself is known for cryptocurrency
undermining and other applications. The paper argues that one should look beyond the currency applications and
investigate the vast usage of blockchain in e-governance. The paper presents a brief framework of distributed ledger
systems in the field of land administration which provides trustworthy record keeping. Our research has concluded that
inclusion of blockchain technology in land administration can overcome the drawbacks of the traditional methods. This
is the platform where one can abolish the problems facing in the land registration, not only in India as well as many
parts of the world. Hence using blockchain Technology to save the land record transaction is the way to create the
immutable records. This technology has added additional features to the platform of land-records.
________________________________________________________________________________
Keywords: Blockchain, e-governance Land registration, Land administration, Manipulation.
cryptographically with a digital firewall
Introduction
generated by a hashing function. Each block is
Today's blockchain is like the steam engine in "chained" to the previous one by referring to its
1698. extraordinary and game-changing. In the hash value. The computers, or nodes, that
past few years, economists thought of connect to the blockchain verify that the record
introducing distributing ledger technology in is valid per the rules of the governing. In 2016
the field of land administration system. Which the Swedish government introduced blockchain
brought revolutionary changes in the well in their administration system, which digitized
profound land registry system, by establishing their record with the decreased paper work
a trustworthy and well-functioning land record Lantmateriet save $106 million per annum.
system. using the descriptions of the current state of
In decades of practicing manual records, which play of various proofs of concept in countries
left the information wide open for fraud and that are piloting a blockchain-based Land
corruption. To overcome the drawbacks in the Administration and by relating these
traditional method which is slow and developments to the practices of well-
cumbersome. blockchain and distributed ledger functioning modern Land Administrations. It
systems have come into the light. blockchain has to be noticed here that developments are
provides a secured record system without any growing at a rapid pace and (blockchain)
need for a trusted third party. while drafting technology is changing almost all the time.
this paper most accurate and precise This paper, therefore, gives a temporary
information has been consulted, which overview of the recent developments and
provided information that insecurity, possibilities(1).
manipulation of records, corruption, double or
multiple registrations are the major problems Land registration
faced by the taxpayers in the traditional Land registration refers to a system whereby a
method. blockchain is based on cryptography government entity records ownership and land-
just like bitcoin which enables its capacities related rights. These Land-records provide
beyond any other technology. evidence of title, facilitate transactions, and
In simple terms, A blockchain is a prevent fraud. According to the World Bank,
continuously growing, distributed, shared 70% of the world’s population lacks access to
ledger of blocks, which are sealed land titles. For the government, records of land
ownership are necessary to collect taxes, apathy towards cross-referencing and

provide services, and set up its territorial verification of data.
authority. The World Bank has shown its 2. Lack of Standardization: - The DILRMP
efforts in improving land registration in several system is dependent upon the data and
countries. In emergent nations the provocation information received from the village level
is twofold. On one hand, securing land government staff. The village-level
property rights is a crucial factor in government maintains a sizeable
encouraging economic development. On the handwritten amount of data with them,
other hand, in developing countries, land which leads to a lack of standardized data,
records are typically kept on paper in a unify complex to retrieve data, and wear-tear of
location(2). handwritten data.
Currently, the process will be computerized, 3. Manipulation: - The registered
and a Land Information system will be information is not standardized due to a
developed to improve estimation and Land lack of teamwork between the various
Title Registration According to the land Act departments who were handling land
1999, three land occupancy rules are identified; records and property ownership data.
the constitutional rights of occupancy, the Further, the process of updating the Land-
customary freedom, and other informal rights. records is lengthy and clumsy. Therefore,
To ensure property rights security and realize the nature of ownership is often not
the benefits from it, governments need to reflected in many of the records, hence
perform three functions that build on each making it strenuous to trace if any peculiar
other, specifically (i) unambiguous definition entry has been made.
and enforcement of property rights; (ii) 4. Lack of Coordination: - There is a lack of
provision of reliable information; and (iii) cost- communication and coordination between
effective management of land-related various sectors such as land records
externalities. management, land registration, judiciary,
taxation, etc. Land registration provides an
The following are the steps for the process of
assent between the two parties on public
land registration:
notice, but the registry is silent on the legal
1. First Investigate the Land Title Exist in validity of the underlying transaction.
Land Registry Therefore, the title to the underlying land
2. Go to the Land and Survey the land is purely tentative.
3. Get the confirmation of land Title 5. Litigations: - The title registration is the
4. Get the recording of the property’s Global open challenge in a court of law, which is
Positioning System (GPS) coordinates. a tentative nature. Currently, there are
5. Next, Enter the collected data and nearly 30 million pending cases, in which
information into the land title database. 70% are related to land and property
6. Issuing the Proofs of property ownership disputes.
to the rightful property owner. 6. Fraud and Black Money: - Property
7. Therefore, the land Registration process fraud is one of the significant problems
has been ended. facing in land records. A report from the
Flaws in Treditional Land Administration Controller and Auditor General (CAG) of
System India describes that in 2015 nearly 1.5 lakh
cases were pending under the Registration
The following are flaws faced in Land Act. Besides, one more major mode of
Records(figure 1): - parking is black money. Black money is
1. Lack of Professional Approach: - The usually dumped through the Benami (false
process of providing land records is name) transactions. These two are the
primarily for land revenue collection. The major flaws that we are facing today in
village-level land revenue accountant and Land-Records.
the land revenue inspector tend to display
Figure 1: Flaws in traditional land administration

Blockchain Methodology(1-3) participant’s computers (acting as nodes).Here,
each block in the blockchain references to the
Blockchain- Blockchain can be defined as a
previous one and also has contents of data, its
shared ledger that stores data (or) information
own hash, and also the hash of the previous
about the transaction in a distributed way.
block (figure 2). Each hash can be compared to
Working- By this definition, we can explain it a fingerprint as it’s unique and can be
easily, that these transactions are performed generated using a string of text using a
within a blockchain between the people, here mathematical function (4-5).
the blockchain is formed by various
Figure 2: Methodology of blockchain

In a blockchain, if any of the block’s hash tool for identifying the change in data, which
value is to be changed, it can make the whole are stored in each block(6).
blockchain invalid, thereby it acts as a great
Figure 3: Depiction of valid blockchain
In figure 3, every hash value matches its previous hash value. So, it’s a valid blockchain.
Figure 4: Representation of invalid blockchain

In figure 4, due to the change in the hash value way one can detect when the document is
in the single block, the whole blockchain altered or missing, implementing this method
becomes invalid(6). in the land administration system which results
However, the joining of any new block is in authenticity and accountability of the land-
allowed in this peer-to-peer network, and that related documents. Blockchain replaces the
block also gets the whole copy of the centralized system with a distributed database
blockchain. Effective hashing and proof-of- that helps protect the central repositories from
work mechanisms with Distributed storage of replication or duplication (1). blockchain
data help to prevent any forgery (or) fraud(4- provides accuracy, availability, and
5). When a document is stored in the authenticity. It’s trustworthy management
blockchain a mathematical module system and administration in the land registry
automatically generates a hash value, which gain a lot of people's trust and support.
makes the identity of the document unique.
Introducing Block Chain To Land
when two documents contain a single hash
Administration
value, they are likely to be neglected. this hash
values make it easy to access the document. Blockchain can be used in different methods to
Most importantly this method is fully secured, process information. each method differs from
whenever an alteration is made to a document one another in privacy and trust. one of the
in the blockchain its hash value also changes simplest ways to use blockchain is by sorting
making it impossible to go unnoticed, by this the hash value of the document (7).
Figure 5: Flow chart showing the blockchain technology
In general, there is the other method to people used complex concepts like off chains
implement blockchain in the land registry here and smart contracts.
Figure 6: Flow chart showing the trust worthiness of blockchain technology
Figure 7: Distributed applications of blockchain technology
Here all the transactional data of the land is data stores that are separate from the
stored in off chains (stored out of blockchain) blockchain itself and can be updated when
which are later transfers into the blockchain transactions occur(4-5).
using smart contracts. smart contracts are user
Conclusion
created business rules such as transfer of
ownership and eligibility of lease. These smart The article under consideration presents in the
contracts present in distributed ledger upcoming year that Blockchain technology will
determines what transaction to be recorded in play an important role in the industry. Now
the blockchain.These also have the ability to there are many systems which is based on
read and write through program interfaces to Blockchain technology which is already in
existence. There are many problems such as
collaboration of brokers or middleman, time would make land registry falsification-free

delays, etc. This is the platform where we can distributed ledger open for anyone to view. It
abolish the problems facing in the land can help by preventing fraudulent to interfere
registration, not only in India as well as Many between a buyer and seller. The operation of a
parts of the world. Hence using blockchain blockchain is a shared responsibility. The
Technology to save the land record transaction suitable governance and encouragement for
is the way to create the immutable records. operation has to be set up. The goal is to make
This technology has added additional features the solution public source to a large range. This
to the platform of land-records. methodology will also need to incorporate with
current systems and procedures, especially
Future Scope
banks and real estate agents, and the Land
Blockchain technology would bring great Registry in order to assure that the service
impact on Land Registration by making it more works with the collaborators involved in the
efficient. Switching to blockchain technology procedures(6).
References
1. Jacob Vos, Christiaan Lemmen, Bert http://stdm.gltn.net/docs/Designing-a-

Beentjes, Blockchain-Based Land Land-Records-System-for-The-Poor.pdf
Administration Feasible, Illusory or A Panacae? 6. Telia company, Chroma Way and Kairos
https://www.researchgate.net/publication/3 Future, The Land Registry in the
16595854_Blockchain- blockchain - testbed, A development
based_Land_Administration_Feasible_Illu project with Lantmäteriet, Landshypotek
sory_or_Panacae Bank, SBAB, march 2017
2. Tien Tuan Anh Dinh, Rui Liu, Meihui 7. Christiaan Lemmen, Jacques Vos and Bert
Zhang, Beng Chin Oo, Untangling Beentjes (2017), Ongoing Development of
Blockchain: A Data Processing View of Land Administration Standards:
Blockchain Systems, IEEE Transactions Blockchain in Transaction Management,
On Knowledge and Data Engineering, EPLJ; 6(3): 478–502.
VOL. 30, NO. 7, JULY 2018 https://doi.org/10.1515/eplj-2017-0016
https://ieeexplore.ieee.org/stamp/stamp.jsp
8. Christidis, K. and Devetsikiotis, M. (
?tp=&arnumber=8246573
2016) Blockchains and Smart Contracts
3. Marten Risius Kai Spohrer, A Blockchain
for the IoTs, IEEE Access, Special section
Research Framework, 14 August 2017, US
on the plethora of Research in IoT.
Government 2017
https://s3.amazonaws.com/academia.edu.d https://www.researchgate.net/profile/Kama
ocuments/58091770/A_blockchain_Resear nashis_Biswas/publication/311716550_Se
ch_Framework_treated.pdf?response- curing_Smart_Cities_Using_Blockchain_
content- Technology/links/5a2682d5a6fdcc8e866be
disposition=inline%3B%20filename%3DA cfb/Securing-Smart-Cities-Using-
blockchain_Research_Framework_treated.pdf Blockchain-Technology.pdf
4. Victoria L. Lemieux, Blockchain and 9. Distributed Ledger Technology: beyond
Distributed Ledgers as Trusted block chain. Report. UK Government
Recordkeeping Systems: An Archival Chief Scientific Adviser, Government
Theoretic Evaluation Framework, Future Office for Science.
Technologies Conference (FTC) 2017.
https://www.researchgate.net/publication/3
17433591
5. “Designing a Land Records System for
The Poor,” (report, United Nations Human
Settlements Programme, 2012),12,
AN IMPACT OF TEACHING STYLE OF LEADERSHIP ON SENIOR SECONDARY

STUDENTS
R. K. Pathak1 and D. Saxena2

1,2
Institute of Education and Research, Mangalayatan University, Aligarh
_______________________________________________________________________________________
ABSTRACT
The teacher, plays an important role to plan and give direction to students’ life. He appears to be in leading position
for the students in view of the content that follows the activities to make themselves as decision maker. His / her works
treated as the impact on styles of leadership in class room teaching. the leadership can affect their psychological
development in their personality enhancement for further activities within these certain objectives.1. If teacher-
students’ relations are based on co-operation and involvement then the students’ level of satisfaction towards the
activity performed. The Analysis of the teacher-student relationship in relation with the way the students feel this
interaction. 2. The exploration of the pupils’ opinions regarding the characteristics of the desired teacher. Research
hypotheses: 3. If teacher-students’ relations are based on co-operation and involvement then the students’ level of
satisfaction towards the activity performed. Research hypotheses: Students get the feeling they are in control of the
activities of leadership that they are about to exist in their personality. If the teacher creates the style of leadership. It
may communicate and transmit the skills, of self-development. The sample has been taken from 200 senior secondary
school students. Conclusions: a teacher sets up his / her relations with the students is the most important concerned
subject like assimilation; suggesting; lifelong training programmes, preparing of modules on pupil-focused learning.
The training of teachers for pupil-focused counselling should be in the moderate way to complete the modules for
personal development and teachers-students based programme for students learning destined to teachers.
________________________________________________________________________________
Keywords: leadership style, teacher-student relationship, student-focused learning
Introduction leading behaviour of a teacher can suggest to

best alternatives to the students. The first is to
1- The authoritative leadership Style
let them decide to resolve the issue of the
It is a style in which a teacher can make all the common things as well and the second is to
leadership activity and take decision for his observe a common criterion of assessment of
organisation. He performs and share his overall students. This democratic quality of a teacher
planning with assessment of criteria for gives students to act in such a way to
performance in groups and individual. These organisation a member of a group and to lead
activities of leadership provide him a process, it.
plans and strategy in the field of education
where he can maintain a position of a leader. 3- The Laissez-faire Leadership Style
The field in which the teacher is engage should This style is a teacher means he is a follower of
be authoritative where he can freely take the passivity. He is a passive leader who limits his
hole management and leadership style to show participation in various activities. He gives
off his activities and personality. This provides students to take all initiatives and establishes
the students to choose the teacher who can friendly attitude with students. He is a man of
prepare them to study with full authority and non-involvement behaviour and cannot takes
embarrassment. division in the mean-time. He hesitates on such
occasions when some offering, suggestion and
2- Democratic Leadership Style
advice relevant to topic has been given to him
In this style the teacher proves him-self as a as a task.
democratic leader who does not participate
neither too much not to less not to less in the 2. Research hypotheses
further activities of teaching. He avoids making If teacher students’ relations are based on co-
decision along but invited the group of operation in enrolment regarding to the aim of
participants to settle a matter of organisation. students’ satisfaction level towards those
The students can learn such opportunity to activities which increase the level of
chose such feature of teaching with an advice understanding the problems of each other.
to sought regarding the fulfilment the task. The Teacher students’ relations are based on
participation and increasement where the Data Analysis

students feel that they are in control of the It is according to the provided the data of
management or the teacher to convey the students of teacher exhibit in awareness and
activities the duties to perform. The teacher communication at the level of class reached up
exhibits the real communication and relating to 60% satisfactory while compare to
skills in his vocation to him self as an ideal overburden of work minimise to 60% and
teacher. This conduct of relations between maximise to 70% through the medium of
teacher and students gives a support to each extend that is necessary to know a good
other that can enhance the satisfactory level in communication needs to be appreciable in their
teaching and other activities. personal requirement. Further activities of class
4-Research Methodology room teaching will be carried on under the
guidance of teacher. It is considerable that the
The research was conducted on the base of
percentage of the students over 50% that the
study about the different style in class room
relation between the teacher and students
teaching. The leadership style in the class room
provide a co-operation in completing data.
performed the various analysis of teacher
While it is mention that a teacher will be
students’ relations, characteristics monitoring
powerful and the students feel the dominant the
and other way to established interaction that
class while the teacher try to work in relation
can influence students’ attitude and behaviour.
of equality. In such condition we found there is
The methodology is on the level of
no co-operation among the members. The
satisfactions that can estimate the capacity and
teacher should use the technique to motivate
pedagogic talent in the students. This process
the students and encouraged them with full
of research method can be conducted through
appreciation. The final analysis of the study
opinion poll enquiries by the help of
where the teacher students’ relation should
questionnaire to average the highest and lowest
according to the criteria over the activities in
extend in students who have essential merits to
which they are to deploy. The certain result
be an ideal teacher.
may be identified from the appreciation.
The following seven questions are different
a) Are the class mates mutually integrated to
variants so the students can answer the best
solve the task?
corresponds to his opinion.
b) What is the percentage of appreciation
Q-1 Do you appreciate that the activity carried
between students and teacher relationship?
on in the class room under teacher
c) The standout views that lead to the
coordination?
following consideration.
Q-2 Are these co-orations among the students
The authoritative teacher those whose work
and teacher can fulfil the task of class room
style is to impose themselves not to cooperate
teaching?
with students. Students should not feel the
Q-3 Do the class room students prove them self
sense of unity thus they offer the positivity to
as a capable of interpersonal awareness and
be more persevering the certain aspect the class
communication?
room teaching.
Q-4 Do the leadership teaching provide the
The democratic style a relay important for the
equal opportunities to all participant.
students and teacher for a good communication
Q-5 How the democratic leadership can change
and establishing a proper understanding among
the environment of teaching class.
them. Therefore, it is their talent to prove in the
Q-6 How are the various activities will be
class room.
helpful for both teacher and students while
The laissez-faire leadership teaching style is
teaching through laissez-faire leadership style.
not appreciable in condition because it left the
Q-7 Are the students of the class mutually
impression form the teacher to be passive on
intergraded and accepted the challenge to solve
sides either be a teacher or an advisor to the
the task together.
students.
Conclusion There for the analysis of the characteristic of

teacher and conduct while establishing the
As a witness in the different context leadership
relation with students is an important element
style. It is predominant in all aspects in all
of his character. It attributes him to be an ideal
analysis of teacher students’ relationship comes
teacher if he has these qualities like good
to feel increase and motivated to participate in
communicator intelligence in working, friendly
future action. It is affirmed that the perception
behaviour and fairness and patience. The
of the level of students may varied while they
portrait of the ideal teacher even he has quality
work individual and while they work in groups.
of leadership and focus vision on students
As for the identification the students felling of
would be an idea teacher and his quality will be
unity and their tendence to appreciate the
explorable.
relations and their development so far in future.
References
1. Bamburg, J. and Andrews, R. (1990). Change. San Francisco: Jossey Bass Fullan,
School goals, principals and achievement, M. (2002). The change leader, Educational
School Effectiveness and School Leadership”, 59(8), 16–20
Improvement, 2(3), 175–191 9. Hall, B. W., Pearson L. C., and Carroll, D.
2. Bateman, T. S. and Organ, D. W. (1983). (1992). Teachers’ long-range teaching
Job satisfaction and the good soldier: The plans: A discriminated analysis, Journal of
relationship between affect and employee Educational Research, 85(4), 221-225
‘citizenship, Academy of Management 10. Hallinger, P. (2000). A review of two
Journal, 26(4), 587-595 decades of research on the principalship
3. Blase, J. and Blaze, J. (1999). Effective using the Principal Instructional
instructional leadership: Teachers’ Management Rating Scale, Paper presented
perspectives on how principals promote at the annual meeting of the American
teaching and learning in schools, Journal of Educational Research Association, Seattle,
Educational Administration, 38(2), 130-141 Washington
4. Bogler, R. (2001). The Influence of 11. Hallinger, P. (2003). Leading educational
Leadership Style on Teacher Job Change: reflections on the practice of
Satisfaction, Educational Administration instructional and transformational
Quarterly, 37(37), 662-683 leadership, Cambridge Journal of
5. Day, C., Harris, A. and Hadfield, M. Education, 33(3), 329-351
(2001). Challenging the orthodoxy of 12. Hallinger, P. and Murphy, J. (1986). The
effective school Leadership, International social context of effective schools,
Journal of Leadership in Education, 4(1), American Journal of Education, 94(3),
39–56 328–355
6. Flores, M. A. (2007). The impact of school 13. Harris, A. and Lambert, L. (2003). Building
culture and leadership on new teachers’ Leadership Capacity for School
learning in the workplace, International Improvement. London: Open University
Journal of Leadership in Education: Theory Press
and Practice, 7(4), 297-318 14. Heller, H., W., Clay, R. & Perkins, C.
7. Ford, M. A. (1992). Motivating humans: (1993). The relationship between teacher
Goals, emotions and personal agency job satisfaction and principal leadership
beliefs. Newbury Park, CA: Sage style Journal of School Leadership, 3(1),
8. Fullan, M. (2001). Leading in a Culture of 74-86

13jay Chotaliya

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

13jay Chotaliya

Uploaded by

Copyright:

Available Formats

Vidyabharati International Interdisciplinary Research Journal ISSN: 2319-4979

ASM: ADAPTIVE SEQUENTIAL MODELLING BASED RECOMMENDATION SYSTEM

S. Singh1* and K. Srivastava2

Introduction recommendations. These filtering methods

Literature Survey Hinton proposed deep learning-based model on

Special Issue on Recent Research Trends in Management, Science and

Special Issue on Recent Research Trends in Management, Science and

Special Issue on Recent Research Trends in Management, Science and

5: Comparison of KNN, SVD, SVD++ and

1. Davidson, James et al. (2010). “The recommendation systems. In: International

Special Issue on Recent Research Trends in Management, Science and

collaborative filtering. IEEE Internet Guangzhong Sun. xDeepFM (2018)

AUTOMATIC SOL-GEL TECHNIQUE BASED ANALYSIS ON OPTICAL PROPERTIES

Introduction optical gadgets over GaN. Moreover, the

Fig. 5 Plot of (ahv)2 vs. hv of ZnO thin films.

Fig. 6 Plot of (ahv)1/2 vs. hv of ZnO thin films.

3. G. Sberveglieri, S. Groppelli, and G. 12. Muthuvinayagam, N. Melikechi, P. D.

CHOICE BASED CREDIT SYSTEM IN INDIA: A CRITICAL REVIEW OF

F. Lalrinzuali and R.P. Vadhera

h) Transferring of Credits 3. CBCS puts the focus on the 'learner', thus

Conclusion present scenario of different universities across

DECEPTIVE SUBLIMINAL STRATEGIES IN ADVERTISING: A SEMIOLOGICAL

Gopakumar AV1, Renjini T2 and Sudheer S S3

Introduction example, audio stimuli may be played below

of the images, and this suppressed information Low-intensity

DEEP LEARNING FOR SENTIMENT ANALYSIS: A REVIEW

A. Save* and N. Shekokar

Introduction they have bought.

Neural Network in sentiment classification. paper claims to have experimented with

results obtained are comparatively higher. The combination of machine learning

neural network applied in series, a combination • Sentiment Analysis of

network and convolution neural network in Fig. 2: Categories of Deep L Learning

Table 1: Analysis of deep learning techniques in sentiment analysis (A)

title must be capitalized except articles,

Table.2. Analysis of deep learning techniques in sentiment analysis (B)

Accuracy 87.6% 93% 91.85% Not mentioned 86.2%

Table.3. Analysis of deep learning techniques in sentiment analysis (C)

Number of Not Mentioned 2 2 2

From the review it can be understood that 3. Conclusion

1. C. Piyaphakdeesakun, N. Facundes, and J. Classification”, IEEE Access, Vol. 8, 2020,

16. S. Anbukkarasi and S. Varadhaganapathy, AAAI Conference on Artificial

OPTIMIZED 2DOF-PIDN CONTROLLER BASED AGC SYSTEM OF A MULTI SOURCE

1. Introduction problem of AGC can be classified into fast

Fig. 1 Block diagram for 2 DOF-PIDN controller

2.2 Buzzard Optimization Algorithm based

Table 1 Optimal control parameters of the for the test system

Fig. 4 (a) Dynamic responses of the frequency deviations of Area-1

Fig. 4 (b) Dynamic responses of the frequency deviations of Area-2

Fig. 4 (c) Dynamic responses of the Tie-line power deviations (actual)

Fig. 4 (d) Dynamic responses of the Tie-line power deviations (error)

Fig. 5 (a) Dynamic responses of the control input deviations of Area-1

Fig. 5 (b) Dynamic responses of the control input deviations of Area-2

Fig. 6 (a) Dynamic responses of change in power generation of Genco-1

Fig. 6 (b) Dynamic responses of change in power generation of Genco-2

Fig. 6 (c) Dynamic responses of change in power generation of Genco-3

Fig. 6 (d) Dynamic responses of change in power generation of Genco-4

6. Conclusion based test system ensures improved ASR

Table.1: Control area, Gencos parameters [3, 10]

OPTIMIZED FREQUENCY AND POWER CONTROL STRATEGY FOR HYBRID

1. Introduction pollution and cost, it is essential to integrate a

obtained as ⃗. The new vector positions for the

responsibilities such as place search, hunting, ⃗ = ⃗. ⃗ − ⃗

system, and the simulation is performed in model. Proposed model performance is