Chapter 2

REVIEW OF RELATED LITERATURE

This chapter includes related literature and related studies that give ideas to the

researcher in the formulation of the study. The reviews found on the literature published

and other studies.

Facial Recognition and Voice Recognition

A facial recognition system is a computer application capable of identifying or

verifying a person from a digital image or a video frame from a video source. One of the

ways to do this is by comparing selected facial features from the image and a facial

database. Face detection is a computer technology being used in a variety of applications

that identifies human faces in digital images. Face detection also refers to the

psychological process by which humans locate and attend to faces in a visual scene. It is

typically used in security systems and can be compared to other biometrics such as

fingerprint or eye iris recognition systems. (Parveen kumar & Doulat singh, 2016).

In this study, face recognition is implemented and used for identifying the user

accessing the system proposed. It capable of identifying the student from video frames

captured by the systems camera. Locate and extract facial features then compare it to a

known facial measurements of the students in the system.

 10

Face Recognition and Detection Techniques

On a study conducted by Parveen kumar and Doulat singh (2016), Approach on

Face Recognition & Detection Techniques, Images are an important form of data and are

used in almost every application. Some applications cannot use images directly due to the

large amount of memory space needed to store these images. One of the most critical

decision points in the design of a face recognition system is the choice of an appropriate

face representation. Effective feature descriptors are expected to convey sufficient

invariant and non-redundant facial information. Motion information is used to find the

moving regions and probable eye region blobs are extracted by thresholding the image.

These blobs reduce the search space for face verification which is done by template

matching. Experimental results for face detection show good performance even across

orientation and pose variation to a certain extent. The face recognition is carried out by

cumulatively summing up the Euclidean distance between the test face images and the

stored database, which shows good discrimination for true and false subjects. As human

face is a dynamic object having high degree of variability in its appearance, that makes

face detection a difficult problem in computer vision. In this field, accuracy and speed of

identification is a main issue. The study concludes that image based approaches is best

among other approach for face recognition. Image-based approaches are having three

methods: Linear subspace methods, neural networks, and statistical approaches.

Face Recognition Algorithms

On a survey paper written by Marqu´es, Ion (2010) Face Recognition Algorithms,

the paper presented the face recognition area, explaining different approaches, methods,
 11

tools and algorithms used since the 60’s. Some algorithms are better, some are less

accurate, some of them are more versatile and others are too computationally costly. The

paper also discusses feature extraction methods for face recognition. The paper also

discusses problems with face recognition like pose and illumination and how they can

possibly be solved using different approaches.

In this study, knowing known face recognition and detection techniques is trivial.

Also taking into account the speed and accuracy of each technique for face recognition

provides the researchers of this study some useful information and approaches to improve

face recognition. The “Face Recognition Algorithms” paper guide the researchers on how

to resolve some of the significant issues on face recognition such as pose and illumination

and also help the researchers of this study to weigh the most effective algorithms to be

used on face recognition. Thus the above mentioned face detection, recognition,

technique, and algorithms provide very useful information in the field of face recognition.

Histogram of Oriented Gradients Algorithm

On a paper by Dalal, Navneet and Triggs, Bill (2005) Histograms of Oriented

Gradients for Human Detection, the paper shows that after reviewing existing edge and

gradient based descriptors, shows that grids of Histograms of Oriented Gradient (HOG)

descriptors signicantly outperform existing feature sets for human detection. The study

evaluates the influence of each stage of the computation on performance, concluding that

Fine-scale gradients, fine orientation binning, relatively coarse spatial binning, and

High-quality local contrast normalization in overlapping descriptor blocks is all important

for good results.

 12

In this study, HOG algorithm is implemented and used for locating and

classifying if there is a human face on an image. HOG method is based on evaluating

well-normalized local histograms of image gradient orientations in a dense grid. Similar

features have seen increasing use over the past decade. The basic idea is that local object

appearance and shape can often be characterized rather well by the distribution of local

intensity gradients or edge directions, even without precise knowledge of the

corresponding gradient or edge positions. In practice this is implemented

by dividing the image window into small spatial regions ("cells"), for each cell

accumulating a local 1-D histogram of gradient directions or edge orientations over the

pixels of the cell. An overview of the feature extraction and object detection chain shown

in Figure 2. The detector window is tiled with a grid of overlapping blocks

in which Histogram of Oriented Gradient feature vectors are extracted. The combined

vectors are fed to a linear SVM for object/non-object classification. The detection

window is scanned across the image at all positions and scales, and conventional non-

maximum suppression is run on the output pyramid to detect object instances, but this

paper concentrates on the feature extraction process.

Figure 2. Overview of the HOG Feature Extraction Chain Face Landmarks

Alignment
 13

Face Landmark Alignment

On the paper by Kazemi, Vahid and Sullivan, Josephine (2014), One Millisecond

Face Alignment with an Ensemble of Regression Trees, the paper presents a general

framework based on gradient boosting for learning an ensemble of regression trees that

optimizes the sum of square error loss and naturally handles missing or partially labelled

data, the paper presents a new faster algorithm for face alignment and face landmark

estimation. The paper addresses the problem of Face Alignment for a single image. It

show how an ensemble of regression trees can be used to estimate the face’s landmark

positions directly from a sparse subset of pixel intensities, achieving super real time

performance with high quality predictions. It described on how an ensemble of regression

trees can be used to regress the location of the face landmarks like the top of the chin, the

outside edge of each eye, the inner edge of each eyebrow, lips, etc. Shown in Figure 3 an

ensemble of randomized regression trees is used to detect 194 landmarks on face from a

single image in a millisecond.

Figure 3. Selected results on the HELEN Dataset

 14

In this study, the algorithm use by Kazemi, Vahid and Sullivan, Josephine (2014)

is implemented to locate face landmarks such as eyes, nose, lips, eyebrows etc. on a

given image and then rotate the images so that these landmarks are centered on the image

frame.

Deep Convolutional Neural Network Algorithms

On a paper written by Google employees Florian Schroff, Dmitry Kalenichenko,

and James Philbin (2015) FaceNet: A Unified Embedding for Face Recognition and

Clustering , the paper present a system called FaceNet, that directly learns a mapping

from face images to a compact Euclidean space where distances directly correspond to a

measure of face similarity. Once this space has been produced, tasks such as face

recognition, verification and clustering can be easily implemented using standard

techniques with FaceNet embeddings as feature vectors. It uses Deep Convolutional

Neural Network algorithms to optimize the embedding itself. The system achieves state-

of-the-art face recognition performance using only 128-bytes per face.

On a paper written by Brandon Amos, Bartosz Ludwiczuk, Mahadev

Satyanarayanan (2016) OpenFace: A general-purpose face recognition library with

mobile applications, the paper presents a face recognition library which implements face

recognition with neural networks from the techniques used by Facebook’s DeepFace and

Google’s FaceNet. Furthermore an article Machine Learning is Fun (Geitgey, 2016). The

article explains on how effective algorithms like Histogram of Oriented Gradients, face

landmark estimation and Deep Neural Networks are for face recognition.
 15

In this study, techniques used by Facebook’s DeepFace and Google’s FaceNet

algorithms have been implemented to extract facial measurements of a human face from

a given image by using OpenFace trained convolutional neural network feature extractor.

It uses a trained neural network that takes 128 facial measurements of a face that is

needed for face recognition. Figure 4 shows the logic flow for face recognition with

neural networks.

Figure 4. Logic Flow for Face Recognition with a Neural Network

Once a face is detected, the systems preprocess each face in the image to create a

normalized and fixed-size input to the neural network. Preprocessing the image

implements algorithm by Kazemi, Vahid and Sullivan, Josephine(2014) normalizing the

faces so that they eyes, nose, and mouth appear at similar locations in each image. Shown

in Figure 5 by using affine transformation. The transformation is based on the large blue

landmarks and the final image is cropped to the boundaries and resized to 96x96 pixels.
 16

Figure 5. OpenFace’s Affine Transformation

The preprocessed images are too high-dimensional for a classifier to take directly

on input. The neural network is used as a feature extractor to produce a low dimensional

representation that characterizes a person’s face. A low-dimensional representation is key

so it can be efficiently used in classifiers or clustering techniques.

Figure 6 illustrates how FaceNet’s training procedure learns to cluster face

representations of the same person. The unit hypersphere is a high-dimensional sphere

such that every point has distance 1 from the origin. Constraining the embedding to the

unit hyper sphere provides a structure to a space that is otherwise unbounded.

Figure 6. Illustration of FaceNet’s Triplet-loss Training Procedure

 17

On a study by Klemen Grm, Vitomir Struc, Anais Artige, et al.(2017), Strengths

and Weaknesses of Deep Learning Models for Face Recognition Against Image

Degradations, the paper studies the effects of different covariates on the verification

performance of four recent deep CNN models using the Labeled Faces in the Wild

(LFW) dataset. The paper specifically investigated the influence of covariates related to:

image quality – blur, JPEG compression, occlusion, noise, image brightness, contrast,

missing pixels; and model characteristics, CNN architecture, color information,

descriptor computation, and analyze their impact on the face verification performance of

AlexNet, VGG-Face, GoogLeNet, and SqueezeNet.Identify the strengths and weaknesses

of the deep learning models, and present key areas for potential future research. The

results indicate that high levels of noise, blur, missing pixels, and brightness have a

detrimental effect on the verification performance of all models, whereas the impact of

contrast changes and compression artifacts is limited. It has been found that the descriptor

computation strategy and color information does not have a significant influence on

performance.

In this study, neural networks models are used to take facial measurements and

thus also present that input image with or that present’s high level of blur, missing pixels,

and brightness thus affect the overall face recognition of capability of the system.

Speech Recognition

Alternatively referred to as speech recognition, voice recognition is a computer

software program or hardware device with the ability to decode the human voice. Voice

recognition is commonly used to operate a device, perform commands, or write without

 18

having to use a keyboard, mouse, or press any buttons. Today, this is done on a computer

with ASR (automatic speech recognition) software programs. Many ASR programs

require the user to "train" the ASR program to recognize their voice so that it can more

accurately convert the speech to text. For example, you could say "open Internet" and the

computer would open the Internet browser. The first ASR device was used in 1952 and

recognized single digits spoken by a user (it was not computer driven). Today, ASR

programs are used in many industries, including healthcare, military (e.g., F-16 fighter

jets), telecommunications, and personal computing (i.e. hands-free computing). In this

study, speech recognition is use to navigate through the system. Voice commands are

given by its users and then it is recognized through speech recognition and is executed by

the system (computerhope.com, 2019).

In this study, speech recognition is implemented and is used for navigation of the

system. It is also used as to operate the system without having to press buttons on the

system.

Voice recognition software on computers requires that analog audio be converted

into digital signals, known as analog-to-digital conversion. For a computer to decipher a

signal, it must have a digital database, or vocabulary, of words or syllables, as well as a

speedy means for comparing this data to signals. The speech patterns are stored on the

hard drive and loaded into memory when the program is run. A comparator checks these

stored patterns against the output of the A/D converter -- an action called pattern

recognition (Rouse, 2018).

 19

In this study, audio signals captured by the microphone is segmented or split into

10-15 milliseconds long. Which is then compared by the system’s acoustic model to get

its features of every segmented signal.

Hidden Markov Model Algorithm

On a paper by Mark Gales and Steve Young (2007), “The Application of Hidden

Markov Models in Speech Recognition”, the paper describes various refinements which

are needed to achieve state-of-the-art performance on continuous speech recognition. It

also discusses Hidden Markov Models (HMMs) could provide a simple and effective

framework for modelling time-varying spectral vector sequences. As a consequence,

almost all present day large vocabulary continuous speech recognition (LVCSR) systems

are based on HMMs.

In this study, Hidden Markov Model algorithm is implemented and primarily used

to recognize speech spoken by the users of the system. The Hidden Markov Model

algorithm implemented in this study is combined with an acoustic model for extracting

statistical features of a given speech. Shown in Figure 7 the basic unit of sound

represented by the acoustic model in the phone.

Figure 7. Architecture of a HMM-based Recognizer in Phones

 20

For example, the word “bat” is composed of three phones /b/ /ae/ /t/. About 40

such phones are required for English. For any given w, the corresponding acoustic model

is synthesized by concatenating phone models to make words as defined by a

pronunciation dictionary. The parameters of these phone models are estimated from

training data consisting of speech waveforms and their orthographic transcriptions. The

language model is typically an N-gram model in which the probability of each word is

conditioned only on its N − 1 predecessors. The N-gram parameters are estimated by

counting N-tuples in appropriate text corpora. The decoder operates by searching through

all possible word sequences using pruning to remove unlikely hypotheses thereby

keeping the search tractable. When the end of the utterance is reached, the most likely

word sequence is output. Alternatively, modern decoders can generate lattices containing

a compact representation of the most likely hypotheses. The feature extraction stage

seeks to provide a compact representation of the speech waveform. This form should

minimize the loss of information that discriminates between words, and provide a good

match with the distributional assumptions made by the acoustic models. For example, if

diagonal covariance Gaussian distributions are used for the state-output distributions then

the features should be designed to be Gaussian and uncorrelated. Feature vectors are

typically computed every 10 ms using an overlapping analysis window of around 25 ms.

One of the simplest and most widely used encoding schemes is based on mel-frequency

cepstral coefficients (MFCCs). These are generated by applying a truncated discrete

cosine transformation (DCT) to a log spectral estimate computed by smoothing an FFT

with around 20 frequency bins distributed non-linearly across the speech spectrum. The

nonlinear frequency scale used is called a mel scale and it approximates the response of
 21

the human ear. The DCT is applied in order to smooth the spectral estimate and

approximately decorrelate the feature elements. After the cosine transform the first

element represents the average of the log-energy of the frequency bins. This is sometimes

replaced by the log-energy of the frame, or removed completely.

CMU SPHINX-4 Speech Recognition System

On a paper by Paul Lamere, Philip Kwok, Evandro Gouvea, Bhiksha Raj ,et al.

(2003), THE CMU SPHINX-4 SPEECH RECOGNITION SYSTEM, the paper has

described the salient features of the architecture of CMU’s latest Sphinx-4 speech

recognition system. Also have described the advantages of the system’s modular design,

its flexibility in the usage of various kinds of acoustic and language representations, and

the features due to the Java platform that it has been developed on. The recognizer

incorporates a completely new algorithm in its search, the Bush-derby algorithm, in

addition to the conventional Viterbi search algorithm. The Bush-derby algorithm enables

the system to include all paths through a unit into the score for the unit. It also facilitates

the use of multiple sources of information, such as audio and visual features, or parallel

feature streams, in a natural way. Shown in Figure 8 is the Sphinx-4 front end module.

The main blocks are the front end, the decoder, and the knowledge base. The decoder

comprises the linguist, the search manager, and the acoustic scorer. The communication

between these modules, as well as communication with an application, is depicted.

 22

Figure 7. Sphinx-4 system architecture.

Figure 8. Sphinx-4 Front End Module

Figure 8 shows a detailed representation of the front-end module. The module

consists of several communicating blocks, each with an input and an output. Each block

has its input linked to the output of its predecessor. When a block is ready for more data,

it reads data from the predecessor, and interprets it to find out if the incoming information

is speech data or a control signal. The control signal might indicate beginning or end of

speech – important for the decoder - or might indicate data dropped or some other

problem. If the incoming data is speech, it is processed and the output is buffered, waiting

for the successor block to request it. The decoder block consists of three modules: search

manager, linguist, and acoustic scorer. The primary function of the search manager is to

construct and search a tree of possibilities for the best hypothesis. The construction of the

search tree is done based on information obtained from the linguist. In addition, the
 23

search manager communicates with the acoustic scorer to obtain acoustic scores for

incoming data. HMM is performed in two ways: depth-first or breadth-first. Depth-first

search is similar to conventional stack decoding. Breadth-first search is performed using

the standard Viterbi algorithm as well as a new algorithm called Bush-derby. The

algorithm works that rather than considering only the highest scoring path for each unit

which surrounded by dotted boxes shown in Figure 9, this algorithm combines all paths

of the unit into a single score. During search, competing units (phoneme, word, grammar

etc.) are each represented by a DIRECTED ACYCLIC GRAPH (DAG). Each dag has a

source and a sink, as illustrated by the examples in figure 9 it shows two-node DAGS for

two competing phonemes AX and AXR. The DAGS associated with the competing word

units CAT and RAT in Figure 3b are more complicated. In Viterbi, each competing unit

is scored by the probability of the (single) best path through it and the unit with

maximum best-path score wins. For example, if the probabilities on the edges were

(0.9,0.02,0.01) and (0.2,0.7,0.6) for the phonemes AX and AXR, respectively, then the

scores would be 0.9 and 0.7 and AX would win. If we used the sum of the probabilities of

all paths rather than the maximum, then the phoneme AXR would come out higher.

Figure 9. Bush-derby Algorithm

 24

In this study, the researchers utilized the CMU Sphinx-4 speech recognition

system for its modularity, flexibility and algorithmic aspects. It uses newer search

strategies, is universal in its acceptance of various kinds of grammars and language

models, types of acoustic models and feature streams. Algorithmic innovations included

in the system design enable it to incorporate multiple information sources in an elegant

manner. The system is entirely developed on the Java™ platform and is highly portable,

flexible, and easier to use with multithreading.

Understanding the CMU Sphinx Speech Recognition System

In a paper “Understanding the CMU Sphinx Speech Recognition System” by

Chun-Feng Liao (2016), the purpose of the project is to find out how an efficient speech

recognition engine can be implemented ,the author examined the source code of Sphinx2

carefully and figured out the role and function of each component. The project found that

the core of Sphinx speech recognition system is composed of three components:

FrontEnd , Knowledge base and Decoder . Frontend receives and processes speech

signals. Knowledge base provides data for decoder and the decoder searches the states

then return the result .A speech recognition system provides some APIs to be

manipulated by voice applications.

In this study, the paper by Chun-Feng Liao helped the researchers understand the

core architecture of cmu sphinx and how the principal components of the system work.
 25

Pocketsphinx

In a paper submitted to the IEEE International Conference on Acoustics Speech

and Signal Processing Proceedings by David Huggins-Daines, Mohit Kumar, Arthur

Chan, et al. (2006), POCKETSPHINX: A FREE, REAL-TIME CONTINUOUS SPEECH

RECOGNITION SYSTEM FOR HAND-HELD DEVICES, the paper attempt to address

the issues of deploying speech recognition applications on embedded or mobile devices.

The most difficult of these is the computational requirements of continuous speech

recognition for a medium to large vocabulary scenario. The need to minimize the size and

power consumption for these devices leads to compromises in their hardware and

operating system software that further restrict their capabilities below what one might

assume from their raw CPU speed. The paper, presents a preliminary case study on the

porting and optimization of CMU SPHINX-II, a popular open source large vocabulary

continuous speech recognition (LVCSR) system, to hand-held devices. The resulting

system operates in an average 0.87 times real-time on a 206MHz device, 8.03 times faster

than the baseline system. To the author’s knowledge, is the first hand-held LVCSR

system available under an open-source license.

In this study, Pocketsphinx modules like the acoustic models and phonetic

dictionary have provided researchers the ability to customize the speech recognition part

of the system proposed in this study. In addition to hardware limitations faced by the

researchers building a speech recognition system from the ground up requires access to

proprietary speech recognition toolkits which are often expensive and usually provided

without source code. Pocketsphinx provided a framework for the researchers to work on.
 26

Kiosk Systems

An electronic kiosk is a computerized or simple help hub in airports, malls, or

other large areas that provide visitors with free information, such as directions or

commercial (computerhope.com, 2018).

In this study, the researchers developed kiosk software with face and speech

recognition that can be installed on a physical hardware and can serve as a student kiosk

where users can view their school records etc.

On a study by Dr. Stephen Jonas, Stephen Megregian, Gary Wenger, et al. (1994),

Implementing a Kiosk-based Campus Information System at Community Colleges: Three

Case Studies, the paper presents case studies of three community colleges that

implemented a campus wide information systems (CWIS) using touchscreen kiosks. The

three colleges cumulatively offer students approximately 20 kiosk encasements on

campus for access to the CWIS. The study concluded that Intouch kiosk system is

reaching a broad spectrum of students: men and women of all ages, ethnic backgrounds,

majors, and academic disciplines. The report titled, 'Who's Using the Kiosks?' notes that

6,480 different students, representing 35 percent of the institution's total spring

enrollment, accessed their personal records (from SCC's student information system) at

least once during the term. Additionally, the statistics showed that kiosks seem to attract

students who are somewhat more dedicated and academically successful than the overall

student population, as indicated by kiosk users having slightly higher grade point

averages and higher credit hours loads than the general student population. An actual

kiosk user distribution breakdown by major closely parallels that of SCC's overall

breakdown by major.
 27

In this study, the study “Implementing a Kiosk-based Campus Information System

at Community Colleges: Three Case Studies” gives insight to the researchers of this study

to further improve and add additional features to the current ISAT U kiosk to

accommodate students with less physical capabilities of accessing their curricular

information in the said kiosk.

Lobby Attendant Software

Lobby Attendant Software[67] virtual front desk kiosk software sold online that

can be installed on any computer and can be placed on building lobby’s the software

contents is customizable depending on the needs. The software can display Building

Directory, Visitor Check-In, Local News & Information, VOIP Calling Text Messaging,

Visitor Badge Printing it can also provide visitors to find their way, search for businesses,

departments and employees, call occupants and other selected services.

The above mentioned software is similar to this study in a sense that they provide

ease of access to information’s need by users. However the software above is target for

businesses and offers much customable user interface and lots of features at a cost.

Information Kiosk System

Information kiosk system by Wizzard Technologies Inc.[68] kiosk software that

enables student to view Class Schedules, Academic Records, Student Tuition fees and

Balances, Update Mobile Number, View Clearance, Courses Offered, Officers’ Directory

or Organization’s Structure, Announcement Message from the President, School

Calendar, Scholarships School Organization and Enrolment Procedure.

 28

The Information kiosk system by Wizzard Technologies Inc. is quite similar to

this study because it is made for schools and for student to view information. However

the software mentioned above the students can update their information on the kiosk and

view school Organization’s Structure etc. In this study the Student kiosk with face and

speech recognition can be further improve and add those additional features so that it can

become a one stop information kiosk for students.

ISO 25010

On a paper by Haslinda, Fahmy, Sukinah, Roslina, Fariha, et al. (2015),

Evaluation of e-Book Applications Using ISO 25010, The paper aims to evaluate the

quality of e-Book applications based on the characteristics set forth by the ISO 25010

standard. The study investigated the quality of e-Book applications in facilitating

classroom learning based on the ISO 25010 standard. A survey of 37 primary schools in

the district of Kemaman in Terengganu Malaysia was carried out involving 200 teachers.

Results indicate that e-Book applications are perceived as usable, reliable, functional, and

efficient in supporting the learning process.

On a paper by Adewole Adewumi and Nicholas Omoregbe (2015), Evaluating

Open Source Software Quality Models against ISO 25010, the paper aims to evaluate

existing open source software quality models against the ISO 25010 standard.

Comparison between ISO 25010 and the other basic quality models showed that the ISO

model was more comprehensive in terms of the number of quality characteristics that it

supported and could serve as a standard.

 29

The quality model is the cornerstone of a product quality evaluation system. The

quality model determines which quality characteristics will be taken into account when

evaluating the properties of a software product.

The quality of a system is the degree to which the system satisfies the stated and

implied needs of its various stakeholders, and thus provides value. Those stakeholders'

needs (functionality, performance, security, maintainability, etc.) are precisely what is

represented in the quality model, which categorizes the product quality into

characteristics and sub-characteristics (iso25000.com, 2018).

Furthermore in this study, ISO 25010 is used to evaluate the system that it

conforms to the standard set by the ISO for computer systems. The system will be

evaluated based on the following quality characteristics standard set by ISO 25010.

Functionality Suitability. This characteristic represents the degree to which a

product or system provides functions that meet stated and implied needs when used under

specified conditions. This characteristic is composed of the following sub characteristics.

Functional completeness, degree to which the set of functions covers all the specified

tasks and user objectives. Functional correctness, degree to which a product or system

provides the correct results with the needed degree of precision. Functional

appropriateness, degree to which the functions facilitate the accomplishment of specified

tasks and objectives.

Performance efficiency. This characteristic represents the performance relative to

the amount of resources used under stated conditions. This characteristic is composed of

the following subcharacteristics. Time behaviour, degree to which the response and

processing times and throughput rates of a product or system, when performing its
 30

functions, meet requirements. Resource utilization, Degree to which the amounts and

types of resources used by a product or system, when performing its functions, meet

requirements. Capacity, Degree to which the maximum limits of a product or system

parameter meet requirements.

Compatibility. Degree to which a product, system or component can exchange

information with other products, systems or components, and/or perform its required

functions, while sharing the same hardware or software environment. This characteristic

is composed of the following sub characteristics. Co-existence, degree to which a product

can perform its required functions efficiently while sharing a common environment and

resources with other products, without detrimental impact on any other product.

Interoperability, degree to which two or more systems, products or components can

exchange information and use the information that has been exchanged.

Usability. Degree to which a product or system can be used by specified users to

achieve specified goals with effectiveness, efficiency and satisfaction in a specified

context of use. This characteristic is composed of the following sub characteristics.

Appropriateness recognizability, degree to which users can recognize whether a product

or system is appropriate for their needs. Learnability, degree to which a product or

system can be used by specified users to achieve specified goals of learning to use the

product or system with effectiveness, efficiency, freedom from risk and satisfaction in a

specified context of use. Operability, degree to which a product or system has attributes

that make it easy to operate and control. User error protection, Degree to which a system

protects users against making errors. User interface aesthetics, Degree to which a user

interface enables pleasing and satisfying interaction for the user. Accessibility, degree to
 31

which a product or system can be used by people with the widest range of characteristics

and capabilities to achieve a specified goal in a specified context of use.

Reliability. Degree to which a system, product or component performs specified

functions under specified conditions for a specified period of time. This characteristic is

composed of the following sub characteristics. Maturity, degree to which a system,

product or component meets needs for reliability under normal operation. Availability,

degree to which a system, product or component is operational and accessible when

required for use. Fault tolerance, degree to which a system, product or component

operates as intended despite the presence of hardware or software faults. Recoverability,

degree to which, in the event of an interruption or a failure, a product or system can

recover the data directly affected and re-establishes the desired state of the system.

Security. Degree to which a product or system protects information and data so

that persons or other products or systems have the degree of data access appropriate to

their types and levels of authorization. This characteristic is composed of the following

sub characteristics. Confidentiality, Degree to which a product or system ensures that data

are accessible only to those authorized to have access. Integrity, Degree to which a

system, product or component prevents unauthorized access to, or modification of,

computer programs or data. Non-repudiation, degree to which actions or events can be

proven to have taken place, so that the events or actions cannot be repudiated later.

Accountability, Degree to which the actions of an entity can be traced uniquely to the

entity. Authenticity, Degree to which the identity of a subject or resource can be proved to

be the one claimed.

 32

Maintainability. This characteristic represents the degree of effectiveness and

efficiency with which a product or system can be modified to improve it, correct it or

adapt it to changes in environment, and in requirements. This characteristic is composed

of the following sub characteristics. Modularity, degree to which a system or computer

program is composed of discrete components such that a change to one component has

minimal impact on other components. Reusability, degree to which an asset can be used

in more than one system, or in building other assets. Analyzability, Degree of

effectiveness and efficiency with which it is possible to assess the impact on a product or

system of an intended change to one or more of its parts, or to diagnose a product for

deficiencies or causes of failures, or to identify parts to be modified. Modifiability,

Degree to which a product or system can be effectively and efficiently modified without

introducing defects or degrading existing product quality. Testability, Degree of

effectiveness and efficiency with which test criteria can be established for a system,

product or component and tests can be performed to determine whether those criteria

have been met.

Portability. Degree of effectiveness and efficiency with which a system, product

or component can be transferred from one hardware, software or other operational or

usage environment to another. This characteristic is composed of the following sub

characteristics. Adaptability, degree to which a product or system can effectively and

efficiently be adapted for different or evolving hardware, software or other operational or

usage environments. Installability, degree of effectiveness and efficiency with which a

product or system can be successfully installed and/or uninstalled in a specified

 33

environment. Replaceability, degree to which a product can replace another specified

software product for the same purpose in the same environment