A Review of Automated Detection Methods for Cyberbullying

Abstract

As we see the cyberspace evolve we also see a directly proportional growth of the people using the cyberspace for communication. As a result, the misuse of the cyberspace has given rise to negative issues such as cyberbullying, which is a form of harassing other people using information technology in a deliberate and continual manner. The detection and prevention of cyberbullying becomes critical for safe and health social media platforms. In this paper, a review of the cyberbullying content in Internet, the categories of cyberbullying, data sources containing cyberbullying data for research, and machine learning techniques for cyberbullying detection are overviewed. The main challenges of the cyberbullying detection are demonstrated, including the lack of multimedia content-based detection and availability of public accessible dataset. Suggestions are provided as the conclusion of the overview.