Detection of Cyber Bullying Using Social Media Network Data

Abstract

The rapid development of online communication and information sharing platforms and the enthusiastic participation of their users has enabled peer-to-peer communication at unprecedented scale and diversity. On the one hand, these communication channels, such as online social networks and news sharing websites, offer myriad opportunities for knowledge sharing and opinion mobilization. On the other hand, they also serve as a fertile domain for an abundance of unfortunate intimidation and hateful aggression and cyberbullying towards individuals targeted because of their identities or expressed opinions. To protect children, it would be beneficial to have technology that can automatically detect and flag cyberbullying. In this research, we explore the use of machine learning in the automated detection of cyberbullying. This dissertation explores the related research and compares several separate machine learning algorithms for this goal. We then conclude with a proposed ensemble approach towards the detection of cyberbullying using a combination of machine learning techniques.