AN INTELLIGENT MACHINE LEARNING FRAMEWORK FOR CYBERBULLYING DETECTION ON SOCIAL MEDIA PLATFORMS

Abstract

The research's goal is to find inappropriate behavior on social media sites using machine learning. Online communication has been significantly enhanced by the expansion of social media. As a consequence, cyberbullying has increased, posing a threat to both physical and mental health. The objective of the project is to create a computer system capable of detecting cyberbullying in extensive social media datasets. Natural Language Processing (NLP) techniques, including sentiment analysis, feature extraction, and text preparation, are employed to identify abusive verbal patterns. Machine learning techniques, including Naïve Bayes, Random Forests, and Support Vector Machines, are employed to construct and evaluate classification models. The proposed method effectively distinguishes between non-bullying and bullying content in order to identify and eliminate detrimental communications. The findings indicate that cyberbullying detection methods are improved by machine learning. The online environment is more complex and responsible.