COMPANY CERTIFICATE

DECLARATION

I, NISARGA B, student of VIII semester B.E in Computer science & Engineering at S J C

Institute of Technology, Chickballapur, hereby declare that the Internship work entitled

“Amazon Product Review Analysis ” has been independently carried out by me under the

supervision of Prof. Harshavardhana Doddamani, Assistant professor, and the coordinator

Narendra Babu C Assistant Professor, submitted in partial fulfillment of the course

requirement for the award of degree in Bachelor of Engineering in Computer Science &

Engineering of Visveswaraya Technological University, Belgavi during the year 2021-2022. I

further declare that the report has not been submitted to any other University for the award of

any other degree.

PLACE: NISARGA B
Date: USN: 1SJ18CS064

i
 ABSTRACT

A sentiment analysis API uses natural language processing (NLP) tasks to not only identify
aspects of the products from the Amazon reviews but also enable brands to look beyond star
ratings. Amazon review data analysis can give insightful customer information that can be
harnessed for product betterment.
We further scrutinized the reviews to analyze the qualitative sentiment beyond the numerical star
rating given. Sometimes people provide higher ratings, and their reviews are bad. The reviews
are written by actual users of the product, which means, what’s written is usually a first-hand
experience. his data is a treasure trove for brands because people can be very vocal about their
experience and give key information about the positives and negatives of a product, including its
delivery and the customer service they received
Another important reason is that even though a product may have received 4 stars, it doesn’t
necessarily mean that the product was good. Digging deeper into the comments, many a time,
shows that there are many negative points that the review mentions even though the person may
have given a higher star rating. And vice versa.

ii
 ACKNOWLEDGEMENT

With reverential pranam, we express my sincere gratitude and salutations to the feet of his
holiness Byravaikya Padmabhushana Sri Sri Sri Dr. Balagangadharanatha Maha Swamiji,
& his holiness Jagadguru Sri Sri Sri Dr. Nirmalanandanatha Swamiji of Sri
Adichunchanagiri Mutt for their unlimited blessings. First and foremost we wish to express my
deep sincere feelings of gratitude to our institution, Sri Jagadguru Chandrashekaranatha
Swamiji Institute of Technology. For providing me an opportunities for completing my
internship work successfully.
I extend deep sense of sincere gratitude to Dr. G T Raju, Principal, S J C Institute of
Technology, Chickballapur, for providing an opportunity to complete the Internship Work.
I extend special in-depth, heartfelt, and sincere gratitude to our HOD Dr. Manujunath
Kumar, Professor and Head of the Department, Computer Science and Engineering, S J C
Institute of Technology, Chickballapur, for her constant support and valuable guidance of the
Internship Work.
I convey our sincere thanks to Internship Internal Guide Prof. Harshavardhana
Doddamani, Assistant Professor, Department of Computer Science and Engineering, S J C
Institute of Technology, for his/her constant support, valuable guidance and suggestions of the
Internship Work.
I am thankful to Internship External Guide Dhanush S, Project Engineer, Compsoft
Technologies, Rajajinagar for providing valuable guidance and encouragement of the
Internship Work.
I also feel immense pleasure to express deep and profound gratitude to our Internship
Coordinator Prof. Narendra Babu, Assistant Professor, Department of Computer Science
and Engineering, S J C Institute of Technology, for his guidance and suggestions of the
Internship Work.
Finally, I would like to thank all faculty members of Department of Computer Science
and Engineering, S J C Institute of Technology, Chickballapur for their support.
I also thank all those who extended their support and co-operation while bringing out this
Internship Report.

Nisarga B (1SJ18CS064)

iii
 CONTENTS
Declaration i
Abstract ii
Acknowledgement iii
Contents iv
List of Figures vii

Chapter No Chapter Title Page No

1 COMPANY PROFILE
1.1 History of the Organization 1
1.1.1 Objectives 1
1.1.2 Operations of the Organization 2
1.2 Major Milestones 2
1.3 Structure of the Organization 3
1.4 Services Offered 4

2 ABOUT THE DEPARTMENT

2.1 Specific Functionalities of the Department 6

3 TASK PERFORMED
9
4 REFLECTION NOTES
4.1 Experience 14
4.2 Technical Outcomes 14
4.2.1 System Requirement Specification
4.3 System Analysis and Design 15
4.3.1 Existing System 15
4.3.2 Disadvantages of the Existing System 16
iv
 4.3.3 Proposed System 16
4.3.4 Advantages of the Proposed System 16
4.4 System Architecture
4.4.1 Data Flow Diagram 17

4.5 Implementation 17
4.5.1 Modules
4.6 Screen Shots 18
5 CONCLUSION
21
BIBLIOGRAPHY
22

v
 LIST OF FIGURES

Figure No. Name of the Figure Page No.

Figure 4.4.1.1 Data Flow Diagram. 17

Figure 4.6.1 Output 1 18
Figure 4.6.2 Output 2 18
Figure 4.6.3 Distribution of Rating 19
Figure 4.6.4 Number of Reviews for Top 20 Brands 19
Figure 4.6.5 Number of Reviews for Top 50 Brands 20

vi
vii
CHAPTER - 1
COMPANY PROFILE

1.1 History of the Organization

Compsoft Technologies is founded by a group of Enterprise Architects having over two
decades of experience in software architecture, design and development of mission critical
systems for some of the Fortune 500 companies. We hire some of the best talents in the
market to deliver quality software on your aggressive milestone dates. Being in the industry
and having hands on experience, we fully understand the entire Software Development Life
Cycle and we only hire resources who meet our high standards. All of our resources go
through our rigorous interview process based on your requirements and we only select the
candidates who not only technically strong but also they are fully dedicated to deliver on your
promise, the success of your organization.

Our resources are expert in designing and developing applications using Agile and Scrum
methodologies. Whatever your software development methodologies may be, our resources
have experience in broad areas and they can pull any project successfully.
We work hard to enhance continuously our reputation for accessibility, professionalism,
performance, and the depth and quality of our long-term consultative relationships with our
clients. We endeavor to be valued as an industry leader in client satisfaction, quality
performance and reputation. All activities will be conducted to the highest ethical and
professional standards.
To help our clients achieve their objectives by serving as their manpower consulting firm.
Compsoft Technologies has one-to-one relationships with a number of clients, helping them
benefit from all of the technologies available to them and build a better solution that exceeds
client's expectation. It is our goal that offers a full range of software, consulting, support,
automation combined with a wide range of technologies that enable clients to consider how
they could achieve their objectives.

1.1.1 Objectives
We are committed to going the extra mile to bring success to the clients consistently. We are
dedicated to delivering the right people, solutions, and services to the clients that they require
to meet their technology challenges and business goals.

1
Amazon Product Review Analysis Company Profile

Delivering the most efficient and the best solution to our clients to every client leveraging
leading technologies & industry best practices.

1.1.2 Operation of the Organization

The race for digital transformation is on. In this globally connected on-demand world with
rapid advancements in internet technologies, businesses worldwide are under constant
pressure to add innovative real-time capabilities to their applications to respond to market
opportunities.

Every business worldwide is building event-driven, real-time applications - from financial

services, transportation, and energy, to retail, healthcare, and Gaming companies.
Our endeavour is to make it easy to develop innovative real-time applications and efficient to
operate them in production.

We have a proven record of building highly scalable, world-class consulting processes that
offer tremendous business advantages to our clients in the form of huge cost-benefits,
definitive results and consistent project deliveries across the globe.

We prominently strive to improve your business by delivering the full range of competencies
including operational performance, developing and applying business strategies to improve
financial reports, defining strategic goals and measure and manage those goals along with
measuring and managing them.

1.2 Major Milestones

Compsoft Technologies is an Information Technology manpower consulting and product
development firm specialized in bringing businesses and technologies together under the
same umbrella. Our in-depth knowledge in developing mission critical systems for Fortune
500 companies has earned us the reputations in the marketplace. Whether you are building
complex systems for your customers or migrating your legacy systems to cutting edge
technologies, our resources are fully trained and equipped with the knowledge required to
perform the job right, from the very first day. We serve on a wide variety of our clients
including Banking, Accounting, Insurance, Healthcare, Retail, Trucking & Transportation.

Dept. of CSE, SJCIT 2 2021-2022

Amazon Product Review Analysis Company Profile

We have proven record of evaluating the best candidates for your requirement and stand by
on the quality throughout the project implementation.

In today’s competitive marketplace, it is important to bring the businesses and technologies

together to deliver on your promise. More than ever, Compsoft Technologies is committed to
deliver on our promise so that you can deliver on yours, the success of your organization.

1.3 Structure of the Organization

CST, core strength lies in our super energetic and gigantic team, forming an excellent blend
of IT minds along with a creative bent. Taking ownership of not only one’s own task but also
creating an enduring & contented customer is what every individual works together with. Our
endeavour is to continuously improve and deliver maximum– enriching products & solutions.
That’s what we mean – doing IT better driving business transformation, digitally – Applying
disruptive technologies for business transformations. CST brings great advantage to business
space by bridging gaps, simplifying businesses and elevating competitive benefits by
providing technology-based business solutions.

Leveraging the enormous talent of our passionate and proven individuals. We are hugely a
customer-centric organization that is bent upon consummating the needs of the customers
beyond their expectations. We successfully host a consortium of experienced
professionals who work in synergy in order to gain an edge over the market. we look at
ourselves as a team where we co-create with them.

Having delivered successful projects we pride ourselves on being a sought-after mobile

application development. Through the years, and have been successfully delivering value to
our customers. We truly believe that our customer's success is our success We don’t look at
ourselves as a vendor for their projects instead. You would be excited to hear some of our
stories and know to what extent we have gone in the interest of the success of our customers.
and we work hard to make that happen. This philosophy and execution have resulted in a
long-term partnership with most of our clients.

Dept. of CSE, SJCIT 3 2021-2022

Amazon Product Review Analysis Company Profile

1.4 Services Offered

It is believed that service and quality is the key to success, Enabling business success driven
by technology. Harnessing the power of technology, we create a measurable difference for
our clients across various industries & multiple geographies.
1 Development.
We develop responsive, functional and super-fast websites. We keep User Experience in
mind while creating websites. A website should load quickly and should be accessible
even on a small view-port and slow internet connection.
2 Branding and Design.
We offer professional Graphic design, Brochure design & Logo design. We are experts
in crafting visual content to convey the right message to the customers. We also design
custom wraps for your products(also known as package designing).
3 Search Engine Optimization
We help you manage your SEO campaign more efficiently and effectively. We help you
gain market share by leveraging our expertise. our holistic approach to identify anything
that may be hurting your traffic or rankings and show you just how to outrank the
competition.
4 Content Writing
We provide content writing services for blogs and product descriptions, our team helps
you generate content to Increase your Brand Recall. We can amplify your marketing
needs & help you reach your potential customers.
5 Research
We equip business leaders with indispensable insights, advice and tools to achieve their
goals, our main area of research is in sentimental analysis, having published multiple
papers on the same, we are in the process of creating a virtual bot that is intended to use
our sentimental analysis data to provide real time replies.
6 Embedded System and IOT
CST works with Consumer Electronics, Lighting, Home Automation, Metering, Sensor-
Technology, Home Appliance and Medical Device companies to help them create smart
and connected products.

Dept. of CSE, SJCIT 4 2021-2022

Amazon Product Review Analysis Company Profile

Through its integrated Embedded and IoT services, Techno soft helps build intelligent &
connected devices that can be remotely monitored and controlled while leveraging edge
and cloud computing for a host of intelligent applications and analytics.
7 Full Stack Web Development
Full stack web development is the practice of working on both the front-end and back-
end of a program. Full Stack is a layer of software or web development consisting of the
front-end and the back-end portions of an application. Front-end is what the users will
see or interact with on your application. Back-end part is what users do not see, such as
application’s logic, database, server, etc. A full-stack web developer is comfortable
working with both back-end and front-end technologies which make a website or
application function properly.

Dept. of CSE, SJCIT 5 2021-2022

CHAPTER – 2
ABOUT THE DEPARTMENT

2.1 Specific Functionalities of the Department

Our department of tech support majorly focused on manage, maintain and repair IT systems.
The Special functionalities include
 Understanding the work to be completed.
 Planning the assigned activities in more detail if needed
 Completing assigned work within the budget, timeline and quality expectations
 Informing the project manager of issues, scope changes, risk and quality concerns
 Proactively communicating status and managing expectation

2.2 Process Adopted

The department aims to first understand the user requirements. Further on, a basic structure of
the product that needs to be built is drawn and understood. Eventually, the technologies that
would best help in developing the product are understood. If the product has database
requirements, the schema and the database design are worked upon. The department believes
in “Think before you code”- the requirements and logics are first understood over a paper and
then are moved to a code form. Agile processes generally promote a disciplined project
management process that encourages frequent inspection and adaptation, a leadership
philosophy that encourages teamwork, self-organization and accountability, a set of
engineering best practices intended to allow for rapid delivery of high-quality software, and a
business approach that aligns development with customer needs and company goals. Agile
development refers to any development process that is aligned with the concepts of the Agile
Manifesto. The Manifesto was developed by a group fourteen leading figures in the software
industry, and reflects their experience of what approaches do and do not work for software
development.
2.3 Testing
Testing was done according to the Corporate Standards. As each component was being built,
Unit testing was performed in order to check if the desired functionality is obtained. Each
component in turn is tested with multiple test cases to verify if it is properly working. These
unit tested components are integrated with the existing built components and then integration
testing is performed. Here again, multiple test cases are run to ensure the newly built

6
Amazon Product Review Analysis About the Department

component runs in co-ordination with the existing components. Unit and Integration testing
are iteratively performed until the complete product is built. Once the complete product is
built, it is again tested against multiple test cases and all the functionalities.
The product could be working fine in the developer’s environment but might not necessarily
work well in all other environments that the users could be using. Hence, the product is also
tested under multiple environments (Various operating systems and devices). At every step, if
a flaw is observed, the component is rebuilt to fix the bugs. This way, testing is done
hierarchically and iteratively.

2.4 Structure of the Department

Figure 2.4.1 Structure of the Department

2.5 Roles and Responsibilities of Individuals

 Since the internship was online, to ensure easy onboarding of interns, the company
had additional individuals who took care of the smooth run of online training.
 Operation and Strategy Head- Ensured there were no difficulties for interns while
onboarding. Best of mentors and doubt clarifying sessions were arranged too.
 Technical Lead- Ensured the technicalities of online training to be smooth. Best
platforms were arranged for our meetings and trainings.

Dept. of CSE, SJCIT 7 2021-2022

Amazon Product Review Analysis About the Department

 Mentors- They have helped us to understand the concepts, gave us tasks to get
practical take a ways and clarified doubts to the best.
 Interns- Worked through the tasks given either individually or in a group

Dept. of CSE, SJCIT 8 2021-2022

CHAPTER – 3
TASK PERFORMED
During the internship I was a part of Sentimental Analysis Research Team. Sentiment
analysis has been discussed in research papers since at least 1954. Early on, companies faced
a major problem of how to respond to customer feedback which was often contradictory. The
introduction of sentiment analysis helped companies to be able to handle this issue by looking
at the general tone of the comments, rather than the specifics, and tailor their responses
accordingly.
A common process works by first identifying the emotions behind each word - whether that's
anger, disgust, fear, joy, sadness, or surprise. This is done through natural language
processing, which analyses the meaning behind each word. This process is then followed by
machine learning and natural language understanding which can work out what the text is
about - what caused the emotions or how they are related to each other.

Sentiment analysis has come a long way since it was first introduced decades ago and it's
know something we use in our day-to-day conversations. From Siri answering questions in a
more humanized way to chatbots being able to understand what we want from them; this field
has grown considerably in recent years but there are still challenges that need solving if it's
ever going to be perfect.

In this project we design an interface for a retailer to register and login. If the retailer is the
existing candidate, then he can directly login. Once the retailer login in, he can view the
review analysis. The reviews are categorised as good, bad and average. Similarly based on
the name of the customer pie chat of gender who had posted the review is also seen.

User Interface:

The user interface developed using python language on anaconda platform. It is achieved by
importing tkinter module. Functions were created for a new register, assigning function to
new register,login page and the password check

On successful login the analysis were displayed to the retailer/user.

At the Backend,

9
Amazon Product Review Analysis Task Performed

The Steps are as follows:

1. Data Scraping : Crawl the amazon review url to extract all required details from it.
We need to take care of the text so as to satisfy the required format, for e.g. tags have
a special meaning to the browser i.e. break read or next line, we need to explicitly
convert each
tag to spaces or else the crawling result will be improper. When working with online
reviews there is always a question in our mind, how can I trust the review. This is not
a problem with amazon reviews, amazon reviewers can up vote or down vote a
review, this collectively is available as helpful count.The datasets were provided by
the company for the analysis

2. Data Cleaning and Processing: The data extracted need to be cleaned so that we get
proper text review on which analysis can be performed. Cleaning of crawled data is
done by removal of all special characters (such as: “:/.,’#$*^&-) in order to retrieve
best results. After cleaning the crawled content copy it into a csv file. The next step is
processing the cleaned data, firstly review is classified as service, feature or product
review. If the review is a feature review then feature extraction is done using POS
Tagging and grammar rule all stated below. After feature extraction the feature
opinion polarization is obtained.

3. All processed output is stored in one csv file for further use.

4. Finally the summarization of sentiments is generated as charts and displayed to the

user as an attractive dashboard.

Algorithm 1.

1. Crawl the amazon review csv to extract all required details from it. Special
care for required format of information must be taken.

2. Cleaning the crawled data. Removal of all special characters (such as :

“:/.,’#$*^&-) must be done in order to retrieve best results. This also saves
our review processing time. Put the crawled content into a csv file.

Dept. of CSE, SJCIT 10 2021-2022

Amazon Product Review Analysis Task Performed

3. Read the csv file for processing, for each review do the following:

i. Perform a service review test where the review is tested for occurrence
of+++++ service words, i.e. if the review length is shorter than 15
words and service words are found in the review the the review is
classified as service review else if the length of review is greater than
15 then more than 2 service must occur in the review for it to be a
service review.

ii. If the review fails for the service test then it is tested for features of a
product (such as camera, microphone and battery) if these exist then
we classify the review as a feature review.

i. For each feature we extract its sentiment from the review using POS
tagging and ruled based extraction (using regular expressions).

ii. Each phrase of sentiment extracted above is then sent to polarizer

that return 1 if the sentiment is positive else -1 which means the
sentiment is negative.

iii. If the review fails the feature test also, then the review is classified as a
product review. A new final csv is generated with the classification
and sentiment of the feature phrases.

4. This csv is then loaded into the database for creating the visualizations by
querying data from the database.

Sentiment Polarity Methodology

The service and the product review’s polarity is the rating the user provides for that review.
The Good Reviews are those with rating 5 stars and 4 stars, Average Reviews are those with
rating 3 stars and Bad Reviews are those with rating 2 stars and 1 star. Finally, when a feature
sentiment is extracted the sentiment phrase is sent to a polarizer method, this method
basically returns +1 if the phrase is a positive sentiment else -1 if the phrase is a negative

Dept. of CSE, SJCIT 10 2021-2022

Amazon Product Review Analysis Task Performed

sentiment. Firstly, the phrases are tested for indirect opinions such as “Battery no better than
iPhone 4s”, the test phrase is tested for certain pre-defined phrases that were found during
manual analysis of reviews. Next if the phrase test fails, the review is tested for the word
“not” if the word not exists then everything after not is polarized meaning every word after
not is tested for whether it is a positive word or a negative word and consecutive words
polarity are added and finally negated, for example “Camera is not good” this phrase is
classified as negative as the word “good” is negated by the word “not”. Lastly if “phrase” and
“not” test fail the test phrase is broken down into words and polarity of each word is found
from a dictionary of sentiment words bifurcated as good and bad words and collective
polarity is considered i.e. if the sum is below 0 the outcome is negative (-1) else outcome is
positive (+1). Rules for feature extraction

The following are some rules that our model uses to extract feature and its sentiment:

 Adjective + Noun

 Noun +Adjective

 Adverb + Noun

 Adverb +Adjective + Noun

 Noun + Adverb + Verb

 Noun + Verb

 Noun + Verb

 Noun +Verb + Noun

 Noun +Determiner + Adjective

 Noun + Verb + Adverb

 Noun + (verb or Adjective or Adverb)

Algorithm 2
Step 1: Load the important libraries

Dept. of CSE, SJCIT 10 2021-2022

Amazon Product Review Analysis Task Performed
Step 2: Import dataset and extract the X variables and Y separately
Step 3: Divide the dataset into train and test
Step 4: Initializing the SVM classifier model
Step 5: Fitting the SVM classifier model
Step 6: Coming up with predictions
Step 7: Evaluating model’s performance

Abar graph was implemented to view the review results. And the pie chart was implemented
to view the gender of the reviewer.

Dept. of CSE, SJCIT 10 2021-2022

 CHAPTER – 4
REFLECTION NOTES

4.1 Experience
As per our experience during the internship, Compsoft Technologies follows a good work
culture and it has friendly employees, starting from the staff level to the management level.
The trainers are well versed in their fields, and they treat everyone equally. There is no
distinguishing between fresher graduates and corporates, and everyone is respected equally.
There is a lot of teamwork followed in every task, be it hard or easy and there is a very calm
and friendly atmosphere always maintained. There is a lot of scope for self-improvement due
to the great communication and support that can be found. Interns have been treated and
taught well and all our doubts and concerns regarding the training or the companies have
been properly answered. All in all, Compsoft Technologies was a great place for a fresher to
start career and also for a corporate to boost his/her career. It has been a great experience to
be an intern in such a reputed organization.

4.2 Technical Outcomes

4.2.1 System Requirements and Specification

Functional Requirements

These are those requirements which deal with what the system should do for users. It
describes the behavior of the system as it relates to systems it relates to systems functionality
and includes the designation of the required functions.

System Features

Under the system features, the super administrator can have a lot of freedom and do a huge
variety of tasks and manage the system very effectively and efficiently to extract the fullest of
the potential and advantage of the Disease Prediction System
1. The super administrator can manage the information along with adding the symptoms
of the patient.

14
Amazon Product Review Analysis Reflection Notes

2. The user can generate reports based on the symptoms that the patent has provided to
the user. He then directs the patients to take the tests to confirm the presence of the
disease.
3. The hardware requirements are the Intel Core Duo 2.0 GHz Processor with 1 GB or
more RAM and 80 GB or more Hard Disk. It also requires 15th CRT or LCD monitor,
Normal or Multimedia keyboard and a compatible mouse.
4. The software requirements consist of Windows (10 or more) operating system and
Python 3.8 programming language.
3.2 Non Functional Requirements

● It is a requirement specification that specifies criteria that can be used to judge the
operations of a system rather than specific behavior. These requirements are also
called quality attributes of a system as these include the majority those metrics that
define the standard and quality for the system, Some of the parameters coming under
this includes Performance, Security etc.
● Performance: Performance term is mainly used to measure the parameters called time
& space. This project uses verify less space and the actions up or operations
performed are done very quickly in fraction of seconds. There is no issue of memory
size out of bounds.
● Security: Security or authorization is one of the major parameters of all computerized
applications. As details are confidential no malicious user must be allowed to operate
on

4.3 System Analysis and Design

4.3.1 Existing System

The existing system allows to take reviews from the consumers. These reviews must be read
and manually evaluated. Since amazon owns millions of customers evaluating each review is
take slots of efforts. Reviews tell you what products and features trending are, what’s in-
demand, what’s no longer relevant, how your products and those of your competitors are
doing, and more.

It’s no wonder that some sellers have been found willing to pay up to $5 per review with no
questions asked. That said, ensuring that you’re analyzing authentic, uncensored reviews is

Dept. of CSE, SJCIT 15 2021-2022

Amazon Product Review Analysis Reflection Notes

key, as 62 percent of consumers say they will not support brands that engage in review
censorship, and consumers are on the lookout for fake reviews.

As a result, a seller can't afford not to have an effective way to monitor authentic customer
opinions, enabling them to stay ahead of the competition, and secure their place on top of
search rankings.

Properly monitoring these customer opinions also allows sellers to identify trends and
patterns over time so they can improve their products.

What makes this process challenging is that there are millions of new product reviews every
month. This high volume makes it impossible for sellers to monitor these customer opinions
manually, without compromising their profit margins or risking missing out on important
trends or patterns over time.

4.3.2 Disadvantages of the Existing System

 It takes lots of effort to evaluate each review that user provides.

 If reviews are evaluated careless then there may be a chance of business shutdown.

 Few reviews might mis lead other customers.

4.3.3 Proposed System

Amazon is one of the largest online vendors in the World. People often gaze over the
products and reviews of the product before buying the product on amazon itself. But the
reviews on amazon are not necessarily of products but a mixture of product of product review
and service review (amazon related or Product Company related). The buyer is misled as the
overall sentiment (rating classification) that amazon gives is a collective one and there is no
bifurcation between a service review and product review. The proposed model satisfactorily
segregates service and product review, in addition to this it also classifies the review as
Feature review if the user talks about some product feature. A featured review is nothing but
a product review, our model also gives sentiment of the text about the product feature. For
example, if the user writes in his review, “the camera for this phone is very good.”, then we
also classify camera feature as positive. We aim to build a system that visualizes the review’s
sentiment in the form of charts.

4.3.3 Advantages of the Proposed System

Dept. of CSE, SJCIT 15 2021-2022

Amazon Product Review Analysis Reflection Notes

 It helps others to decide on whether to buy a product or not.

 It helps the sellers to scale up his/her sales and get feedbacks from user to
improve the quality of the product.
 It helps you because once you have a good Amazon ranking, you will get
requests to review products on Amazon from different companies. Though the
benefits are not very impressive (they usually send you the product samples, but
it shall surely be helpful to you in expanding your learning curve and contacts).
 You can earn some fame as well if your ranking on Amazon becomes less than
500 in your country.

4.4 System Architecture

4.4.1 Data Flow Diagram

The DFD takes an input-process-output view of a system i.e., data objects flow into the
software, are transformed by processing elements, and resultant data objects flow out of
the software. The dataset contains reviews. The product reviews are obtained. Then the
product reviews are fed to algorithm .Thus ratings are obtained

Model

Dataset Data Pre-processing Algorithm Output

Figure 4.4.1.1 Data Flow Diagram

4.4 Implementation

4.4.1 MODULES

Authorization: The analysis must be done by the retailer to maintain the confidentiality.
Therefore, only authorized users must be able to login to analyze the review.

Data Preprocessing: The data extracted need to be cleaned so that we get proper text
review on which analysis can be performed. Cleaning of crawled data is done by removal

Dept. of CSE, SJCIT 15 2021-2022

Amazon Product Review Analysis Reflection Notes

of all special characters (such as: “:/.,’#$*^&-) in order to retrieve best results. After
cleaning the crawled content copy it into a csv file.

Algorithms: Here we use 2 algorithms.

 The first one is used to segregate the product reviews and the service reviews.

 This algorithm also stores the service review in the form of csv file.

 The second algorithm is support vector algorithm. This algorithm classifies the
reviews as Good, average and bad.

 Using this review the quality of the product can be enhanced

4.6 Screen Shots

Figure-4.6.1 Output 1

Figure-4.6.2 Output 2

Dept. of CSE, SJCIT 15 2021-2022

Amazon Product Review Analysis Reflection Notes

Figure-4.6.3 Distribution of Rating

Figure-4.6.4 Number of Reviews for Top 20 Brands

Dept. of CSE, SJCIT 15 2021-2022

Amazon Product Review Analysis Reflection Notes

Figure-4.6.5 Number of Reviews for Top 50 Products

Dept. of CSE, SJCIT 15 2021-2022

CHAPTER – 5
CONCLUSION
In conclusion, with this study, I tried to show how sentiment analysis works by applying it on
Amazon review data. The system is accurate enough for the test case of reviews on amazon.
For sentiment analysis I have designed my own methodology that integrates existing
sentiment analysis approaches. Classification of reviews along with sentimental analysis
increased the accuracy of the system which in turn provides accurate reviews to the user.

21
 BIBLIOGRAPHY

[1] tweepy.API — Twitter API v1.1 Reference — tweepy 3.10.0 documentation. (n.d.).
Tweepy.Org. https://docs.tweepy.org/en/latest/api.html.
[2] Part 3: The RabbitMQ Management Interface - CloudAMQP. (n.d.). RabbitMQ.
https://www.cloudamqp.com/blog/part3-rabbitmq-for-beginners_the-
management interface.html.
[3] Tutree. (n.d.). Tutree. https://tutree.com/
[4] The Python Tutorial — Python 3.9.4 documentation. (n.d.). Python Tutorial.
https://docs.python.org/3/tutorial/.
[5] Tutorial: Get started with Go - The Go Programming Language. (n.d.). Golang Tutorial.
https://golang.org/doc/tutorial/getting-started
[6] Martin, R. C. (2020, April 7). How to write clean code? Lessons learnt from “The Clean
Code” — Robert C. Martin. Clean Code. https://medium.com/mindorks/how-to write-clean-
code-lessons-learnt-from-the-clean-code-robert-c-martin-9ffc7aef870