PROJECT TITLE
PROJECT PHASE II REPORT
Submitted by
Student Name 1
(Register number)
Student Name 2
(Register number)
Student Name 3
(Register number)
Student Name 4
(Register number)
Under the guidance of
Guide name
Assistant Professor (SG)
in partial fulfillment for the award of the degree of
BACHELOR OF TECHNOLOGY
in
COMPUTER SCIENCE AND ENGINEERING
HINDUSTAN INSTITUTE OF TECHNOLOGY AND SCIENCE
CHENNAI - 603 103
APRIL 2026
BONAFIDE CERTIFICATE
Certified that this project report Cyberbullying Detection Using Deep Learning is
the bonafide work of Student Name 1 (Register number), Student Name 2
(Register number), Student Name 3 (Register number) and Student Name 4
(Register number) who carried out the project work Phase-II (ECY51807) under
my supervision during the academic year 2025-2026.
Name of the guide, Dr. P. Mohankumar,
Supervisor, Professor & Head,
Assistant Professor (SG), Department of Department of Intelligent Systems and
Intelligent Systems and Cyber Security Cyber Security
INTERNAL EXAMINER EXTERNAL EXAMINER
Name:
Name:
Designation:
Designation:
Project Viva - voce conducted on
ACKNOWLEDGEMENT
First and foremost, we would like to thank ALMIGHTY who has provided us the strength
to do justice to our work and contribute our best to it.
We wish to express our deep sense of gratitude from the bottom of our heart to our guide
------------------, Assistant Professor (SG), Department of Intelligent Systems and Cyber
Security, for her motivating discussions, overwhelming suggestions, invaluable
supervision, and exemplary guidance throughout this project work.
We would like to extend our heartfelt gratitude to Dr. J. Thangakumar, Professor &
Associate Dean, School of Computing Sciences for his ingenious encouragement. We
like to express our sincere gratitude to Dr P Mohankumar, Professor & Head,
Department of Intelligent Systems and Cyber Security for his valuable suggestions and
support in successfully completing the project.
We wish to thank our Project guide Dr / Mr /Ms XYX and Panel members for keeping
our project in the right track. We would like to thank all the teaching, technical and non-
technical staff of Department of Computer Science and Engineering for their courteous
assistance.
We thank the management of HINDUSTAN INSTITUTE OF TECHNOLOGY AND
SCIENCE for providing us the necessary facilities and support required for the successful
completion of the project.
As a final word, we would like to thank each and every individual who have been a source
of support and encouragement and helped us to achieve our goal and complete our project
work successful.
ABSTRACT
Cyberbullying has become a prevalent issue with the rise of social media platforms,
creating a need for effective detection systems to combat its negative effects on
individuals and society. This project focuses on utilizing deep learning techniques to
develop an automated system that detects instances of cyberbullying in online text data.
By leveraging the power of Natural Language Processing (NLP) and neural networks,
specifically models such as Convolutional Neural Networks (CNNs), Long Short-Term
Memory (LSTM) networks, and Transformers, the system aims to identify harmful
language patterns, offensive content, and abusive behavior in real-time, providing a timely
response to mitigate its impact. The proposed solution involves preprocessing social media
data, including filtering, tokenization, and embedding techniques such as word
embeddings or transformer-based embeddings to represent the textual content. Deep
learning models are then trained on labeled datasets of online communications to classify
text as cyberbullying or non-cyberbullying. The architecture and training methods will
focus on maximizing the detection accuracy while minimizing false positives and
negatives, ensuring the system's reliability across diverse linguistic and cultural contexts.
Additionally, this project explores the effectiveness of combining deep learning models
with sentiment analysis and keyword-based approaches to enhance detection capabilities.
The cyberbullying detection model's performance will be evaluated using standard metrics
such as precision, recall, F1 score, and accuracy on benchmark datasets. Continuous fine-
tuning and domain-specific training will allow the system to adapt to evolving online
language trends. Ultimately, this project aims to contribute to the growing field of ethical
AI and its application in online safety, offering a scalable and effective solution for real-
world deployment in social media platforms and online communities.
TABLE OF CONTENTS
Table of Figures i
Table of Abbreviations ii
CHAPTER 1 – INTRODUCTION 1
1.1 Overview of the project 1
1.2 Problem Statement 2
1.3 Scope & Motivation for the Project
CHAPTER 2 - LITERATURE REVIEW 3
2.1 Cyberbullying Detection Techniques 3
2.2 Deep Learning Models 4
2.3 Summary 5
CHAPTER 3 – SYSTEM DESIGN 6
3.1 Architecture Diagram 6
3.2 Module Description 9
3.3 Link between Phase I and Phase II
3.3 Pseudocode
3.4 Summary
CHAPTER 4 - EXPERIMENTAL SETUP AND RESULTS 10
4.1 Experimental Setup 10
4.2 Dataset, AI Tools 11
4.3 Result Analysis 17
4.4 Performance Evaluation and Test Case 18
CHAPTER 5 – CONCLUSION 19
5.1 Conclusion 19
5.2 Future Expansion 20
REFERENCES 22
LIST OF APPENDICES
25
APPENDIX A 26
APPENDIX B 27
APPENDIX C 28
APPENDIX D 29
APPENDIX E
TABLE OF FIGURES
FIGURE NO. TITLE PAGE NO.
3.1
Architecture Diagram 6
4.2.1 Output Part 1 11
4.2.2 Output Part 2 12
4.2.3 Output Part 3 12
4.2.4 Output Part 4 13
4.2.6 Confusion Matrix 14
4.2.7 Keywords 15
4.2.8 Model Loss 15
4.2.9 Model Accuracy 16
4.2.10 Accuracy Rate comparison 16
TABLE OF ABBREVIATIONS
[Link] ABBREVIATION DESCRIPTION
1 JSON JavaScript Object Notation
2 HTML Hyper Text Markup Language
5
CHAPTER 1
1.1 OVERVIEW OF THE PROJECT
1.2 PROBLEM STATEMENT
1.3 SCOPE & MOTIVATION FOR THE PROJECT
CHAPTER 2
LITERATURE REVIEW
2.1 CYBERBULLYING DETECTION TECHNIQUES
2.2 DEEP LEARNING MODELS
2.3 SUMMARY
CHAPTER 3
SYSTEM DESIGN
3.1ARCHITECTURE DIAGRAM
Figure 3.1 Architecture Diagram
3.2 MODULE DESCRIPTION
3.3 LINK BETWEEN PHASE I AND PHASE II
3.4 PSEUDOCODE
3.5 SUMMARY
CHAPTER 4
EXPERIMENTAL SETUP AND RESULTS
4.1 EXPERIMENTAL SETUP
(Hardware/ Software requirements)
4.2 DATASET, AI TOOLS
Dataset details, URL of data, sample data, AI tools used
4.3 RESULT ANALYSIS
(Screenshots, Plots and Graphs)
4.4 PERFORMANCE EVALUATION
Performance metrics, equations, tables
4.5 TEST CASE
4.6 SUMMARY
CHAPTER 5
CONCLUSION
5.1 CONCLUSION
5.2 FUTURE EXPANSION
REFERENCES
Formatted in Google Scholar APA Style
APPENDIX A
SAMPLE CODE
import pandas as pd import numpy as np import
tensorflow as tf from
[Link] import
Tokenizer
from [Link] import pad_sequences from
sklearn.model_selection import train_test_split
from [Link] import accuracy_score, classification_report, confusion_matrix,
ConfusionMatrixDisplay import [Link] as plt import
seaborn as sns from wordcloud import WordCloud
# Load the dataset
df = pd.read_excel(r'C:\Users\Jesh\Downloads\youtube_parsed_dataset.xlsx')
# Ensure all comments are strings df['comment'] =
df['comment'].astype(str) # Split the data into training
and testing sets X = df['comment']
y = df['label'].astype(int) # Ensure labels are integers
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) #
Create a tokenizer to split the comments into words
tokenizer = Tokenizer(num_words=5000)
tokenizer.fit_on_texts(X_train)
# Convert the text to sequences
X_train_sequences = tokenizer.texts_to_sequences(X_train) X_test_sequences =
tokenizer.texts_to_sequences(X_test)
# Pad the sequences to have the same length max_length = 200
padded_X_train = pad_sequences(X_train_sequences, maxlen=max_length) padded_X_test =
pad_sequences(X_test_sequences, maxlen=max_length)
APPENDIX B
TEAM DETAILS
INDIVIDUAL CONTRIBUTION:
Team Member 1:
Name: Bavesheya Prabhu (22112078)
Role in project: Handle data collection, cleaning, and preprocessing (e.g., handling missing
data, converting text to sequences), Perform exploratory data analysis, including
visualizations and word clouds, Split the dataset into training and testing sets, ensuring
balance and preparing it for model input.
Team Member 2:
Name: Kavudamalya R (22112090)
Role in project: Evaluate the model with metrics like accuracy, precision, recall, and confusion
matrix analysis, Create visualizations (e.g., confusion matrix, accuracy/loss plots),
Develop the real-time user feedback mechanism and implement the feedback loop for
continuous improvement.
Team Member 3:
Name: Jeshma S J (22112106)
Role in project:
APPENDIX C
Publication/Patent detail
-Acceptance mail, Presented Certificate, Patent Filed
APPENDIX D
Plagiarism Report & AI content Report
APPENDIX E
Contribution to SDG