Abusive Language Detection with Graph Convolutional Networks

TL;DR

This paper introduces a novel graph convolutional network approach that models both community structure and user language behavior to improve abusive language detection on Twitter.

Contribution

It is the first to combine community structure and linguistic behavior in GCNs for abusive language detection, advancing the state of the art.

Findings

Graph-based modeling improves detection accuracy

Captures both community and linguistic features

Significantly outperforms previous methods

Abstract

Abuse on the Internet represents a significant societal problem of our time. Previous research on automated abusive language detection in Twitter has shown that community-based profiling of users is a promising technique for this task. However, existing approaches only capture shallow properties of online communities by modeling follower-following relationships. In contrast, working with graph convolutional networks (GCNs), we present the first approach that captures not only the structure of online communities but also the linguistic behavior of the users within them. We show that such a heterogeneous graph-structured modeling of communities significantly advances the current state of the art in abusive language detection.