DETECTING CYBERBULLYING USING NATURAL LANGUAGE PROCESSING

ABSTRACT The rapid expansion of social media and digital communication platforms has led to an alarming rise in cyberbullying, which negatively impacts individuals' mental and emotional well-being. Traditional approaches to addressing cyberbullying, such as manual moderation and user reporting, often fall short in detecting and mitigating harmful online interactions in real time. This project presents an automated cyberbullying detection system leveraging Natural Language Processing (NLP) and Optical Character Recognition (OCR) to analyse both textual and image-based content. The system processes user-provided text and extracts textual data from images to classify content as either cyberbullying or non-cyberbullying using machine learning algorithms. By integrating deep learning models such as Long Short-Term Memory (LSTM) networks and Convolutional Neural Networks (CNN), the system enhances accuracy in identifying harmful linguistic patterns. The implementation of this detection framework aims to provide an efficient and scalable solution to online safety challenges. Future work will explore multilingual capabilities and real-time detection enhancements to further improve the system's effectiveness across diverse social media platforms.