Machine learning approach for detection of cyber-aggressive comments by peers on social media network

The fast growing use of social networking sites among the teens have made them vulnerable to get exposed to bullying. Cyberbullying is the use of computers and mobiles for bullying activities. Comments containing abusive words effect psychology of teens and demoralizes them. In this paper we have devised methods to detect cyberbullying using supervised learning techniques. We present two new hypotheses for feature extraction to detect offensive comments directed towards peers which are perceived more negatively and result in cyberbullying. Our initial experiments show that using features from our hypotheses in addition to traditional feature extraction techniques like TF-IDF and N-gram increases the accuracy of the system.