CYBER BULLYING DETECTION USING MACHINE LEARNING

Abstract—As the number of people using social media has been increasing exponentially, cyberbullying has evolved as a kind of bullying that occurs through electronic messages. Bullying has been a part of society for as long as anybody can remember. Machine learning can be used to recognize the linguistic characteristics of bullies and so construct a model that can detect cyberbullying automatically. We analyze the strategies to detect   offensive words on social sentences, paragraphs while distinguishing it from general profanity in this paper, which includes a thorough assessment of some previous research on cyberbullying detection tools. Using s classification methods and a manually annotated open-source dataset, we want to develop a interface to detect offensive word identifications. This study presents a complete and structured overview of automatic offensive words detection as well as a systematic comparison of a few of its existing methodologies as  an insightful evaluation of some published research on cyberbullying detection .