Detecting Early Signs of Cyberbullying in Social Media
Nowadays, the amount of users{'} activities on online social media is growing dramatically. These online environments provide excellent opportunities for communication and knowledge sharing. However, some people misuse them to harass and bully others online, a phenomenon called cyberbullying. Due to its harmful effects on people, especially youth, it is imperative to detect cyberbullying as early as possible before it causes irreparable damages to victims. Most of the relevant available resources are not explicitly designed to detect cyberbullying, but related content, such as hate speech and abusive language. In this paper, we propose a new approach to create a corpus suited for cyberbullying detection. We also investigate the possibility of designing a framework to monitor the streams of users{'} online messages and detects the signs of cyberbullying as early as possible.
PDF Abstract
