Published Paper Details:

Social media was a common place for abusive communication in the modern day. More than 80% of online social networks have abusive or vulgar speech on their user profiles, according to a recent survey report. Cyberbullying is the act of threatening or harassing another user via the use of inappropriate, abusive, or vulgar online social media posts. The major purpose of putting these kinds of unpleasant statements on user walls is to harass teenagers, preteens, and other kids. I created the present programme to limit vulgar communication on online social media since, up until now, no application has been able to stop this cyber material from spreading in social networks online. In this suggested application, our major goal is to provide a novel representation learning approach to address the issue of recognising and preventing abusive statements in online chat. Here, we attempt to categorise the misused and legitimate text messages using well-known machine learning methods, including Porter Stemming Algorithm and Support Vector Machine (SVM). The NLT Package (Natural Language Toolkit), which divides the entire message into pieces and then assigns tokens for each and every unique word, is known as Porter Stemming. Here, we divide online bullying into five types, including "HATE, VULGAR, OFFENSIVE, SEX, and VIOLENCE," that are found in the literature.