Understanding Toxic Interaction Across User and Video Clusters in Social Video Platforms

2025-12-10

AI Summary

This study investigates the structural drivers of toxic interactions between users and video clusters on the social video platform Bilibili. We propose a user-video co-clustering framework that jointly models interaction behaviors, textual toxicity/sentiment/length, and content metadata. By constructing an interaction matrix, applying standardized PCA for dimensionality reduction, and performing K-means clustering on both user and video sides, we enable group-level attribution of toxicity. Our key contribution is the first unified, structured analytical paradigm integrating user and video embeddings—moving beyond isolated individual-level analysis. Results reveal that highly exposed video clusters concentrate toxic expressions, warranting prioritized intervention; conversely, user clusters characterized by longer comments and higher comment-to-view ratios exhibit significantly lower toxicity, empirically validating the mitigating effect of reasoned discourse on toxicity propagation. These findings provide interpretable, actionable group-level profiles to inform platform-scale toxicity governance.

Technology Category

• Computer Vision: Video Understanding & Activity Analysis
• Machine Learning: Clustering
• Natural Language Processing: Safety and Robustness

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Web Mining and Content Analysis: Normalization, clustering, classification, and summarization of Web text
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Social video platforms shape how people access information, while recommendation systems can narrow exposure and increase the risk of toxic interaction. Previous research has often examined text or users in isolation, overlooking the structural context in which such toxic interactions occur. Without considering who interacts with whom and around what content, it is difficult to explain why negative expressions cluster within particular communities. To address this issue, this study focuses on the Chinese social video platform Bilibili, incorporating video-level information as the environment for user expression, modeling users and videos in an interaction matrix. After normalization and dimensionality reduction, we perform separate clustering on both sides of the video-user interaction matrix with K-means. Cluster assignments facilitate comparisons of user behavior, including message length, posting frequency, and source (barrage and comment), as well as textual features such as sentiment and toxicity, and video attributes defined by uploaders. Such a clustering approach integrates structural ties with content signals to identify stable groups of videos and users. We find clear stratification in interaction style (message length, comment ratio) across user clusters, while sentiment and toxicity differences are weak or inconsistent across video clusters. Across video clusters, viewing volume exhibits a clear hierarchy, with higher exposure groups concentrating more toxic expressions. For such a group, platforms should require timely intervention during periods of rapid growth. Across user clusters, comment ratio and message length form distinct hierarchies, and several clusters with longer and comment-oriented messages exhibit lower toxicity. For such groups, platforms should strengthen mechanisms that sustain rational dialogue and encourage engagement across topics.

Problem

• Examines toxic interaction clustering in social video platforms
• Models user-video interactions to identify structural and content patterns
• Proposes platform interventions based on exposure and user behavior clusters

Innovation

• Clustering users and videos in interaction matrix
• Integrating structural ties with content signals
• Identifying stable groups for targeted platform interventions