Unmasking Cyberbullies on Social Media Platforms Using Machine Learning

Cyberbullying refers to an aggressive act or behavior with the intention of hurting someone through electronic communication tools. It involves insulting and sending threatening messages to the victim. Cyberbullying incidents are increasing worldwide, posing challenges in combating this form of victimization. Due to the vast amount of online data, tracking cyberbullying is difficult, making detection crucial. Several studies have utilized text mining techniques to analyze and detect cyberbullying. These studies involve classifying conversations or posts using supervised learning, labeling them with N-grams, and applying TF-IDF weighting. In the current system, comments on social media platforms are manually gathered and tagged, then processed using various binary and multiclass classification algorithms. The proposed technique offers practical insights into cyberbullying detection by utilizing deep learning-based models to comprehensively assess cyberbullying across multiple social media platforms and diverse subjects. To address the enormous volume of streaming texts from these platforms, an automatic cyberbullying detection system is being developed. The cluster and discriminant analysis stage analyze the text input to recognize abusive communications. The primary goal of this project is to empower victims by generating thorough reports that can be sent to the appropriate cyber cell for action.