Detection of Cyberbullying Using Machine Learning and Deep Learning Algorithms

IEEE BASE PAPER ABSTRACT:

Use of digital technologies lead to the development of cyberbullying and social media has become a major source for it compared to mobile phones, platforms such as gaming and messaging. Cyberbullying can take several forms that includes sexual remarks, threats, hate mails and posting false things about someone which can be seen and read by millions of people. Compared to traditional bullying, cyberbullying has a longer lasting effect on the victim which can affect them physically or emotionally or mentally or in all the forms. Number of suicides due to cyberbullying has increased in recent years and India is one among the four countries that has more number of cases in cyberbullying. Prevention of cyberbullying has become mandatory in universities and schools due to rising cases since 2015. This paper aims to detect cyberbullying comments automatically using Machine learning and Deep learning techniques. Metrics such as accuracy, precision, recall and F1-score used to evaluate the model performance. It is found that Gated Recurrent Unit, a deep learning technique outperformed all the other techniques which are considered in this paper with an accuracy of 95.47%.

PROJECT OUTPUT VIDEO:

ALGORITHM / MODEL USED:

LSTM architecture.

OUR PROPOSED ABSTRACT:

The widespread use of social media platforms has led to an increase in incidents of cyberbullying, which can have severe psychological and emotional effects on victims. The aim of this project is to develop a deep learning model to detect instances of cyberbullying in social media posts using Long Short-Term Memory (LSTM) networks.

The existing research is directed towards mature languages and highlights a huge gap in newly embraced resource poor languages. This proposed system uses Long Short Term Memory model (LSTM), a deep learning approach, for detecting and preventing cyberbulling actions. The project consists of several modules, including data collection, data preprocessing, model development, and evaluation.

The dataset used in this project was obtained from a popular dataset repository called Kaggle and was preprocessed to remove irrelevant data and convert the text into a numerical format suitable for input into an LSTM model. The LSTM model was trained on the preprocessed data, and its performance was evaluated using various metrics, including accuracy, precision, recall, and F1 score.

The model achieved an accuracy of 95.6% on the test set, indicating that it can effectively detect instances of cyberbullying in social media posts. The trained model was saved and can be used to make predictions on new, unseen data. This project has the potential to be used as a tool to help prevent cyberbullying and provide support to victims.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : Pentium i3 Processor.
Hard Disk : 500 GB.
Monitor : 15’’ LED
Input Devices : Keyboard, Mouse
Ram : 4 GB

SOFTWARE REQUIREMENTS:

Operating System : Windows 10 / 11.
Coding Language : Python 3.8.
Web Framework : Flask.
Frontend : HTML, CSS, JavaScript.

REFERENCE:

Apoorva K G, D Uma, “Detection of Cyberbullying Using Machine Learning and Deep Learning Algorithms”, 2022 2nd Asian Conference on Innovation in Technology (ASIANCON), IEEE Conference, 2022.

Python IEEE Projects

Detection of Cyberbullying Using Machine Learning and Deep Learning Algorithms