Abstract
Cyberbullying has emerged as a pressing concern in various social media platforms, including but not limited to Twitter, Instagram, and Facebook, inflicting both immediate and long-term psychological effects on victims. To combat this pervasive issue, research has sought to build and refine automated systems for cyberbullying detection. This study presents a review of 10 recent AI-powered cyberbullying detection applications, encompassing primarily transformer-based models, their variants and ensemble models. A consolidated framework for designing an effective cyberbullying detection system is also addressed in this paper. It highlights the flow of key components and can serve as a template to ease the design of problem-specific customized systems. Furthermore, AI-powered cyberbullying detection technology has also been widely applied to the education field. Several prevention and intervention applications are outlined and introduced, along with their features and possible drawbacks. Feedback and suggestions from users are also summarized, facilitating the exploration of future research directions.
Author Information
Chun Fai Carlin Chu, The Hang Seng University of Hong Kong, Hong Kong
Hei Nok Charlotte Choy, University of Toronto, Canada
Yee Nim Sarah Kam, The University of Hong Kong, Hong Kong
Comments
Powered by WP LinkPress