Tackling Online Abuse: A Survey of Automated Abuse Detection Methods

TLDR

Online abuse—harassment, racism, personal attacks—affects millions of users, causes lasting psychological harm, and has spurred extensive NLP research to detect it. This survey compiles and evaluates all proposed abuse‑detection methods to provide a foundation for future research. The authors review available datasets, analyze computational techniques and their pros and cons, and discuss emerging trends, remaining challenges, and ethical and explainability guidelines.

Abstract

Abuse on the Internet represents an important societal problem of our time. Millions of Internet users face harassment, racism, personal attacks, and other types of abuse on online platforms. The psychological effects of such abuse on individuals can be profound and lasting. Consequently, over the past few years, there has been a substantial research effort towards automated abuse detection in the field of natural language processing (NLP). In this paper, we present a comprehensive survey of the methods that have been proposed to date, thus providing a platform for further development of this area. We describe the existing datasets and review the computational approaches to abuse detection, analyzing their strengths and limitations. We discuss the main trends that emerge, highlight the challenges that remain, outline possible solutions, and propose guidelines for ethics and explainability