IRJET- Prevention of Cyber Bullying Using Machine Learning Approach (original) (raw)

The online interaction among people happens mostly using social media. There are recent developments in social media. It has challenges to the research community. The challenge is to analyze the online interactions among people. There are several social networking sites where people can share their views on a particular topic. The recent research reveals that on average 20 to 40 % of all teenagers have been victimized because of online social networking sites. In this paper, we mainly focus on the particular form of cyberbullying. This form is nothing but a form of cyber victimization. This can be prevented by adequate detection of such harmful messages. As there is a massive load on the web information, there should be an intelligent system to detect such cyberbullying. The system should identify the potential risk automatically. In this paper, we represent the construction and annotation of a corpus. The fine-grained annotations in cyberbullying such as text categories, insults, abusive words involve cyberbullying. The dataset has the construction of curse words and abusing words. The identification and intimation are done. We present the proof of concept experiments on automatic identification. The finegrained annotations are used for the identification of categories of cyberbullying.