A Study of Cyber Hate on Twitter with Implications for Social Media Governance Strategies

TL;DR

This study investigates cyber hate on Twitter, analyzing user responses and employing machine learning to identify counter-speech, aiming to inform social media governance strategies for mitigating harmful effects.

Contribution

It introduces a mixed methods approach combining qualitative analysis, statistical modeling, and machine learning for understanding and countering cyber hate on Twitter.

Findings

Insights into user interaction patterns with cyber hate posts

A statistical model explaining response volume

Machine learning methods for identifying counter-speech

Abstract

This paper explores ways in which the harmful effects of cyber hate may be mitigated through mechanisms for enhancing the self governance of new digital spaces. We report findings from a mixed methods study of responses to cyber hate posts, which aimed to: (i) understand how people interact in this context by undertaking qualitative interaction analysis and developing a statistical model to explain the volume of responses to cyber hate posted to Twitter, and (ii) explore use of machine learning techniques to assist in identifying cyber hate counter-speech.