SWE2: SubWord Enriched and Significant Word Emphasized Framework for Hate Speech Detection

TL;DR

The paper introduces SWE2, a novel hate speech detection framework that leverages word and sub-word information, demonstrating high accuracy and robustness against adversarial attacks in social media content.

Contribution

SWE2 is a new hate speech detection model that combines word-level semantics with sub-word knowledge, improving accuracy and robustness over existing methods.

Findings

Achieves 0.975 accuracy and 0.953 macro F1 without adversarial attack.

Maintains high performance with 0.967 accuracy under 50% message manipulation.

Outperforms 7 state-of-the-art baselines in experiments.

Abstract

Hate speech detection on online social networks has become one of the emerging hot topics in recent years. With the broad spread and fast propagation speed across online social networks, hate speech makes significant impacts on society by increasing prejudice and hurting people. Therefore, there are aroused attention and concern from both industry and academia. In this paper, we address the hate speech problem and propose a novel hate speech detection framework called SWE2, which only relies on the content of messages and automatically identifies hate speech. In particular, our framework exploits both word-level semantic information and sub-word knowledge. It is intuitively persuasive and also practically performs well under a situation with/without character-level adversarial attack. Experimental results show that our proposed model achieves 0.975 accuracy and 0.953 macro F1, outperforming 7 state-of-the-art baselines under no adversarial attack. Our model robustly and significantly performed well under extreme adversarial attack (manipulation of 50% messages), achieving 0.967 accuracy and 0.934 macro F1.