Measuring Online Hate on 4chan Using Pre-Trained Deep Learning Models

2025-03-30

IEEE Transactions on Technology and Society

AI Summary

This paper addresses the prevalence and complexity of hate speech on the unmoderated, anonymous platform 4chan’s /pol/ board. We propose the first fine-grained, multi-class framework jointly quantifying hate content (e.g., racial, gendered, religious) and toxic behaviors (e.g., identity attacks, threats). Methodologically, we integrate RoBERTa with Detoxify to build a multi-label classification model and augment it with LDA topic modeling to uncover dynamic semantic patterns in hate expression. Our key contribution is the first three-dimensional, synergistic analysis—across hate categories, toxicity intensity, and thematic evolution—conducted entirely within a fully unregulated, anonymous online environment. Experimental results show that 11.20% of posts in our dataset contain at least one hate class, confirming high heterogeneity and blurred categorical boundaries. The framework significantly enhances interpretability and modeling fidelity for unstructured hate discourse.

Technology Category

• Natural Language Processing: Safety and Robustness
• Machine Learning: Multimodal Learning
• Application Domains: Humanities & Computational Social Science

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Responsible Web: Mitigating online harms, including harassment, spam, and fake reviews

Abstract

Online hate speech can harmfully impact individuals and groups, specifically on non-moderated platforms such as 4chan where users can post anonymous content. This work focuses on analysing and measuring the prevalence of online hate on 4chan's politically incorrect board (/pol/) using state-of-the-art Natural Language Processing (NLP) models, specifically transformer-based models such as RoBERTa and Detoxify. By leveraging these advanced models, we provide an in-depth analysis of hate speech dynamics and quantify the extent of online hate non-moderated platforms. The study advances understanding through multi-class classification of hate speech (racism, sexism, religion, etc.), while also incorporating the classification of toxic content (e.g., identity attacks and threats) and a further topic modelling analysis. The results show that 11.20% of this dataset is identified as containing hate in different categories. These evaluations show that online hate is manifested in various forms, confirming the complicated and volatile nature of detection in the wild.

Problem

• Measure online hate prevalence on 4chan using NLP models
• Analyze multi-class hate speech and toxic content dynamics
• Quantify hate speech categories and topic patterns on unmoderated platforms

Innovation

• Uses RoBERTa and Detoxify for hate speech detection
• Applies multi-class classification for hate categories
• Incorporates topic modeling for deeper content analysis