SENTIMENTAL ANALYSIS: ENHANCING PRODUCT

DEVELOPMENT
A MINOR PROJECT REPORT

Submitted by

Abhishek Sharma (2101030600037)

Sandeep Bhagat (2101030600039)

Ujas Rathod (2101030600045)

Preet Singha (2101030600031)

BACHELOR OF TECHNOLOGY

in
Computer Engineering with Artificial Intelligence and Machine Learning

College of Technology
Silver Oak College of Engineering and Technology

Silver Oak University, Ahmedabad

November,2024
 Silver Oak College of Engineering and Technology
Opp. Bhagwat Vidhyapith, S.G. Highway, Ahmedabad-382481

CERTIFICATE
This is to certify that the Minor Project Report submitted along with the
Project entitled Sentimental Analysis: Enhancing Product Development
has been carried out by Abhishek Sharma, Sandeep Bhagat, Ujas Rathod,
Preet Singha under my guidance in partial fulfillment for the Bachelor of
Technology in Computer Engineering with Artificial Intelligence and
Machine Learning, 7th Semester of Silver Oak University, Ahmedabad
during the academic year 2024-25.

A/Prof. Sanket Shah

Internal Guide Head of the Department
 Silver Oak College of Engineering and Technology
Opp. Bhagwat Vidhyapith, S.G. Highway, Ahmedabad-382481

DECLARATION
We hereby declare that the Minor Project Report submitted along with the
Project entitled Sentimental Analysis: Enhancing Product Development
submitted in partial fulfillment for the Bachelor of Technology in Computer
Engineering with Artificial Intelligence and Machine Learning at Silver
Oak University, Ahmedabad, is a Bonafide record of original project work
carried out by me/us at Silver Oak College of Engineering and Technology
under the supervision of A/Prof. Sanket Shah and that no part of this report
has been directly copied from any students’ reports or taken from any other
source, without providing due reference

Name of Student Sign of Student

___________________ ________________
 ACKONWLEDGEMENT

We are heartily thankful to Our supervisor, A/Prof. Sanket Shah, whose encouragement,
supervision and support from the preliminary to the concluding level enabled me to develop
an understanding of the subject. At the end, we offer my regards and blessings to all of
Those who supported us in any respect during the completion of the project and to our
college for providing a resources and materials.

We would like to extend my gratitude to head of Computer Engineering with Artificial

Intelligence and Machine Learning Department, Silver Oak college of Engineering and
Technology, Ahmedabad, for his continuous encouragement and motivation.

Last but not the least we would like to mention here that we are greatly indebted to each
and everybody who has been associated with our project at any stage but whose name does
not find a place in this acknowledgement.

Yours Sincerely,

Abhishek Sharma (2101030600039)

Sandeep Bhagat (2101030600037)

Ujas Rathod (2101030600045)

Preet Singha (2101030600031)

Silver Oak University i Silver Oak College of Engineering & Technology

 ABSTRACT

Sentiment analysis, or opinion mining, is a natural language processing (NLP)

technique that evaluates and determines the emotional tone or sentiment
expressed in text data. This process involves analysing text to identify and
categorize the sentiment as positive, negative, or neutral, and sometimes more
nuanced emotions like joy, anger, sadness, etc. Sentiment analysis has become
increasingly important in various fields, including business, social media,
customer service, and market research.

Sentiment analysis finds applications in diverse domains, such as social media

monitoring, customer feedback analysis, product reviews, brand reputation
management, and financial market prediction.

Silver Oak University ii Silver Oak College of Engineering & Technology

 LIST OF FIGURES

Figure No. Title Page No.

Fig 1.1 Timeline Gantt Chart 4

Fig 3.1 Data-Flow Diagram 7

Fig 3.2 Use-Case Diagram 7

Fig 3.3 Different Classification Approach 8

Fig 4.3.1 Input Block 9

Fig 4.3.2 Review Block 9

Fig 4.3.3 Database Block 9

Silver Oak University iii Silver Oak College of Engineering & Technology
 LIST OF TABLES

Table No. Title Page No.

Table 5.1 Classification Report 10

Silver Oak University iv Silver Oak College of Engineering & Technology

 TABLE OF CONTENT

Acknowledgement i

Abstract ii

List of Figures iii

List of Tables iv

Table of Content v

Chapter 1 Introduction 1

1.1 Introduction 1

1.2 Summary 1

1.3 Purpose 1

1.4 Objective 2

1.5 Scope 2

1.5.1 Document-Level Sentimental Analysis 2

1.5.2 Sentence-Level Sentimental Analysis 2

1.5.3 Aspect-Level Sentimental Analysis 2

1.5.4 Sub-Sentence-Level Sentimental Analysis 2

1.6 Technical and Literature Review 3

1.6.1 Research Papers 3

1.6.2 Books 3

1.6.3 Online Resources 3

1.7 Timeline Chart 4

Silver Oak University v Silver Oak College of Engineering & Technology

 Chapter 2 System Analysis 5

2.1 Study about Current System 5

2.2 Problem with Current System 6

2.3 Requirement of New System 6

Chapter 3 System Design 7

Chapter 4 Implementation 9

4.1 Implementation Environment 9

4.2 Modules 9

Chapter 5 Testing 10

5.1 Testing Plan/Strategy 10

5.2 Test Results 10

Chapter 6 Conclusion 11

References 12

Silver Oak University vi Silver Oak College of Engineering & Technology

 Introduction

CHAPTER-1 INTRODUCTION

1.1 INTRODUCTION

Product-based sentiment analysis is a specialized form of sentiment analysis that

focuses on evaluating customer opinions, feedback, and emotions surrounding
specific products or product features. As businesses increasingly rely on customer
feedback from online reviews, social media, and surveys, sentiment analysis has
emerged as a crucial tool for extracting actionable insights from vast amounts of
unstructured data. By using advanced natural language processing (NLP) and
machine learning (ML) techniques, product-based sentiment analysis can
automatically categorize customer opinions at scale, providing businesses with
valuable data to enhance product offerings, refine marketing strategies, and improve
customer satisfaction.

1.2 SUMMARY

Sentiment analysis is a technique used to automatically analyse text data and

determine the sentiment expressed within it. It's particularly useful for analysing
product reviews, as it helps businesses understand customer feedback, identify areas
for improvement, and make data-driven decisions. The process involves collecting
and cleaning review data, extracting relevant features, applying machine learning
or rule-based techniques to classify reviews as positive, negative, or neutral, and
finally analysing the results to gain valuable insights. By effectively utilizing
sentiment analysis, businesses can leverage customer feedback to make informed
decisions and drive growth.

1.3 PURPOSE

The purpose of sentiment analysis on product reviews is to understand customer

sentiment and gain valuable insights to improve products, services, and overall
customer experience. It helps businesses:

• Understand Customer Feedback

• Identify Areas for Improvement

Silver Oak University 1 Silver Oak College of Engineering & Technology

 Introduction

• Make Data-Driven Decisions

• Enhance Customer Experience
• Gain a Competitive Advantage

1.4 OBJECTIVE

This system helps businesses understand customer sentiments at a granular level,

enabling them to:

• Improve Product Quality

• Enhance Customer Satisfaction
• Drive Data-Driven Product Strategy
• Optimize Marketing and Branding Efforts

1.5 SCOPE

The scope of sentiment analysis on product reviews can be quite broad, encompassing
various levels of granularity and complexity:
1.5.1 Document-Level Sentiment Analysis:
• Determines the overall sentiment of an entire review (positive, negative, or
neutral).
• Provides a high-level understanding of customer feedback.

1.5.2 Sentence-Level Sentiment Analysis:

• Analyses the sentiment of individual sentences within a review.
• Identifies specific aspects of the product or service that customers like or
dislike
.
1.5.3 Aspect-Level Sentiment Analysis:
• Identifies specific product features or aspects mentioned in the review and
determines the sentiment associated with each aspect.
• Provides detailed insights into customer opinions on specific product
features.
1.5.4 Sub-Sentence Level Sentiment Analysis:
• Analyses the sentiment of phrases or clauses within a sentence.
• Captures nuances and subtle expressions of sentiment.

Silver Oak University 2 Silver Oak College of Engineering & Technology

 Introduction

1.6 TECHNICAL AND LITERATURE REVIEW

Here are some key literature sources for sentiment analysis in product reviews:
1.6.1 Research Papers:
• Sentiment Analysis of Product Reviews: This paper provides a
comprehensive overview of sentiment analysis techniques applied to
product reviews. It discusses various approaches, including lexicon-based,
machine learning-based, and deep learning-based methods.
• Deep Learning for Sentiment Analysis of Product Reviews: This paper
explores the application of deep learning techniques, such as Recurrent
Neural Networks (RNNs) and Convolutional Neural Networks (CNNs), for
sentiment analysis.
1.6.2 Books:
• Natural Language Processing with Python: This book covers a wide
range of NLP techniques, including sentiment analysis. It provides
practical examples and code implementations using Python libraries like
NLTK and spaCy.
1.6.3 Online Resources:
• ResearchGate: This platform hosts a large number of research papers on
sentiment analysis, including many related to product reviews.
• Google Scholar: This search engine can be used to find academic papers
and articles on sentiment analysis.

Silver Oak University 3 Silver Oak College of Engineering & Technology

 Introduction

1.7 TIMELINE CHART

Fig 1.1 Timeline Gantt Chat

Silver Oak University 4 Silver Oak College of Engineering & Technology

 System Analysis

CHAPTER-2 SYSTEM ANALYSIS

System analysis for sentiment analysis involves a comprehensive examination of the entire
sentiment analysis system to identify its components, requirements, and objectives. Here is
a breakdown of the key elements of system analysis for sentiment analysis

2.1 STUDY ABOUT CURRENT SYSTEM

The study of the current system of sentiment analysis involves analysing the
existing methods, approaches, and technologies that are being used in sentiment
analysis as of my last knowledge
1. Techniques and Algorithms:
• Investigate the various sentiment analysis techniques and algorithms
currently in use. These include rule-based approaches, machine
learning models, deep learning models, and lexicon-based methods.
2. Preprocessing Methods:
• Examine how text data is pre-processed before sentiment analysis. This
may involve tasks such as tokenization, stop word removal, stemming,
and lemmatization.
3. Feature Extraction:
• Understand the methods for feature extraction, such as TFIDF,
embeddings (Word2Vec, GloVe), and context aware embeddings
BERT.
4. Machine Learning Models:
• Explore the machine learning models commonly employed for
analysis, including Support Vector Machines (SVM), Naive Bayes,
Random Forest, and neural networks (LSTM, CNN).
5. Deep Learning Models:
• Investigate the growing role of deep learning models, such as
transformers (e.g., BERT, GPT-3), in achieving state-of the-art results
sentiment analysis.
6. Datasets:
• Analyse the datasets used for training and testing sentiment analysis
models. This could include well-known datasets like IMDB movie
reviews, Twitter sentiment data, or domain specific datasets.
7. Tools and Libraries:
• Identify the popular sentiment analysis tools and libraries used by
researchers and practitioners, such as NLTK, spaCy, scikit-learn,
Hugging Face Transformers, and TextBlob.

Silver Oak University 5 Silver Oak College of Engineering & Technology

 System Analysis

8. Real-time Analysis:
• Explore systems and technologies that support real-time sentiment
analysis for applications like social media monitoring and customer
service.

2.2 PROBLEM IN CURRENT SYSTEM

sentiment analysis has made significant progress, but it still faces several challenges
and issues in its current existing systems. These problems include:
1. Context Understanding: Sentiment analysis often struggles to fully
understand the context in which text is written. For example, sarcasm and
irony can lead to misinterpretation, as the literal meaning may differ from
the intended sentiment.
2. Ambiguity: The same word or phrase can have different meanings and
sentiments in different contexts. For example, the word "cool" can mean
"stylish" or "temperature" depending on the context.
3. Negation Handling: Sentiment analysis systems may not effectively handle
negations, where words like "not" or "but" reverse the sentiment. For
instance, "not bad" is positive, but a simplistic system may interpret it as
negative.
4. Emotion Detection: Sentiment analysis primarily focuses on positive,
negative, or neutral sentiment, but it often fails to capture the complexity of
emotions like joy, anger, fear, or sadness.
5. Sensitivity to Data and Training Bias: Sentiment analysis models can be
sensitive to the training data and may exhibit bias, reflecting the bias present
in the data used for training. This can lead to unfair or inaccurate results.

2.3 REQUIREMENT OF NEW SYSTEM

When defining the requirements for a new system of sentiment analysis, it's
essential to consider both functional and non-functional requirements. Functional
requirements describe the specific features and capabilities of the system, while
non-functional requirements focus on the qualities and characteristics that the
system should possess.

Silver Oak University 6 Silver Oak College of Engineering & Technology

 System Design

CHAPTER-3 SYSYTEM DESIGN

Fig 3.1 Data-Flow

• DFDs offer straightforward visuals that illustrate the touch points and opportunities
for efficiencies or improvement.
• As visual representations of the system, they make it easier to understand and refine
business processes, keeping them as simple and effective as possible.

Fig 3.2 Use-Case

Silver Oak University 7 Silver Oak College of Engineering & Technology

 System Design

A use case diagram is a fundamental tool in systems engineering and software

development that illustrates how users interact with a system. By mapping out
various use cases, these diagrams offer a clear visual representation of functional
requirements and user behaviours.

Fig 3.3 Different Classification Approach

Two classification approaches of sentiment polarity, machine learning (top), and

deep learning (bottom). Part of Speech (POS); Named Entity Recognition (NER);
Term Frequency-Inverse Document Frequency (TF-IDF).

Silver Oak University 8 Silver Oak College of Engineering & Technology

 Implementation

CHAPTER-4 IMPLEMENTATION

4.1 Implementation Environment

This model is developed to increase the productivity of the product and
know the customers insides of the product they need.
In our model we will collect the reviews of different products and reach out
to more customers and increase business small companies

4.2 Modules
1. Input Block: It takes input from the user to analysis their reviews.

Fig 4.2.1 Input Block

2. Review Block: Here we can see the review result of the user’s input.

Fig 4.2.2 Review Block

3. Database Block: Here we store inputs from the user’s and their type
(Positive/Negative/Natural)

Fig 4.2.3 Database Block

Silver Oak University 9 Silver Oak College of Engineering & Technology

 Testing

CHAPTER-5 TESTING

5.1 TESTING PLAIN/STRATEGY

When we are stuck in the code and did not get desire output, we started debugging
of data and trying to solve errors. And where we stuck our instructor guide us in
the problems as well as errors.

5.2 TEST RESULTS

Sr Model Accuracy Precision Recall F1-score support

No.

1 Gaussian NB 0.884375 0.870665 0.884375 0.873766 320.000000

2 Decision Tree 0.893750 0.886741 0.893750 0.889591 320.000000

3 K Neighbours 0.903125 0.893504 0.903125 0.885766 320.000000

4 Random Forest 0.918750 0.925631 0.918750 0.900040 320.000000

5 Logistic 0.906250 0.915293 0.906250 0.879316 320.000000

Regression

6 Ada Boost 0.831250 0.809528 0.831250 0.819205 320.000000

7 Gradient 0.928125 0.927697 0.928125 0.915752 320.000000

Boosting

Table 5.2 Classification Report

Silver Oak University 10 Silver Oak College of Engineering & Technology

 Conclusion

CHAPTER-6 CONCLUSION

Sentiment analysis has proven to be a valuable tool for businesses to gain insights from
customer feedback. By automatically analysing product reviews, businesses can identify
trends, pinpoint areas for improvement, and make data-driven decisions.

While significant strides have been made in sentiment analysis, challenges such as handling
sarcasm, irony, and complex language nuances persist. Ongoing research and
advancements in natural language processing techniques will continue to improve the
accuracy and effectiveness of sentiment analysis.

By leveraging sentiment analysis, businesses can:

• Enhance Customer Experience: Address customer concerns and improve product

quality.

• Optimize Marketing Strategies: Target specific customer segments and tailor

marketing campaigns.

• Gain a Competitive Edge: Stay ahead of competitors by understanding market

trends and customer preferences.

As technology continues to evolve, sentiment analysis will play an increasingly important

role in helping businesses harness the power of customer feedback to drive growth and
success.

Silver Oak University 11 Silver Oak College of Engineering & Technology

 REFERENCES

1. Najma Sultana, Pintu Kumar, Monika Rani Patra, Sourabh Chandra4 and S.K.
Safikul Alam (2019) “SENTIMENT ANALYSIS FOR PRODUCT REVIEW' -
ICTACT JOURNAL ON SOFT COMPUTING, APRIL 2019, VOLUME: 09,
ISSUE: 03
2. Fang, X., Zhan, J. (2015) “Sentiment analysis using product review data” -
Journal of Big Data 2, 5
3. M. Hu and B. Liu, (2005) “Mining and Summarizing Customer Reviews” -
Proceedings of 10th ACM International Conference on Knowledge Discovery
and Data Mining, pp. 166-177, 2005
4. Aurangzeb Khan, Baharum Baharudin and Khairullah Khan, (2011) “Sentiment
Classification from Online Customer Reviews using Lexical Contextual
Sentence Structure”, - Proceedings of International Conference on Software
Engineering and Computer Systems, pp. 317-331.
5. Richa Sharma, Shweta Nigam and Rekha Jain, 2014,"Mining of Product
Reviews At Aspect Level", International Journal in Foundations of Computer
Science & Technology, Vol.4, No.3.
6. A. Collomb, C. Costea, D. Joyeux, O. Hasan and L. Brunie, “A Study and
Comparison of Sentiment Analysis Methods for Reputation Evaluation”,
Available at: https://liris.cnrs.fr/Documents/Liris-6508.pdf.
7. Mika V. Mantyla, Daniel Graziotin and Miikka Kuutila, (2018) “The Evolution
of Sentiment Analysis-A Review of Research Topics” - Computer Science
Review, Vol. 27, No. 1, pp. 16-32.
8. F. Benamara, C. Cesarano and D. Reforgiato, (2006) “Sentiment Analysis:
Adjectives and Adverbs are better than Adjectives Alone”, - Proceedings of
International Conference on Weblogs and Social Media, pp. 1-7.
9. Andrew L. Maas, Raymond E. Daly, Peter T. Pham, Dan Huang, Andrew Y. Ng
and Christopher Potts, (2011) “Learning Word Vectors for Sentiment Analysis”,
- Proceedings of 49th Annual Meeting of the Association for Computational
Linguistics, pp. 1-7.
10. Understanding Sentiment Analysis: What It Is and Why It’s Used, Available at:
https://www.brandwatch.com/blog/understandingsentiment-analysis/

Silver Oak University 12 Silver Oak College of Engineering & Technology