Evading Overlapping Community Detection via Proxy Node Injection

2025-09-25

AI Summary

This paper addresses the Community Member Hiding (CMH) problem in overlapping community detection—i.e., evading privacy inference based on graph analysis by minimally perturbing the graph structure so that a target node is excluded from its original overlapping community memberships. We formally define this problem for the first time and propose a privacy-preserving framework integrating edge modification and proxy node injection. To jointly optimize perturbation strategies under structural fidelity constraints, we introduce a novel deep reinforcement learning approach tailored to resist mainstream overlapping community detection algorithms. Extensive experiments on multiple real-world social network datasets demonstrate that our method significantly outperforms existing baselines in stealthiness, effectiveness, and computational efficiency. To the best of our knowledge, this work provides the first systematic solution for graph-structural privacy protection in overlapping community settings.

Technology Category

• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Machine Learning: Privacy
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Security and Privacy: Privacy-enhancing technologies
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems

Abstract

Protecting privacy in social graphs requires preventing sensitive information, such as community affiliations, from being inferred by graph analysis, without substantially altering the graph topology. We address this through the problem of emph{community membership hiding} (CMH), which seeks edge modifications that cause a target node to exit its original community, regardless of the detection algorithm employed. Prior work has focused on non-overlapping community detection, where trivial strategies often suffice, but real-world graphs are better modeled by overlapping communities, where such strategies fail. To the best of our knowledge, we are the first to formalize and address CMH in this setting. In this work, we propose a deep reinforcement learning (DRL) approach that learns effective modification policies, including the use of proxy nodes, while preserving graph structure. Experiments on real-world datasets show that our method significantly outperforms existing baselines in both effectiveness and efficiency, offering a principled tool for privacy-preserving graph modification with overlapping communities.

Problem

• Preventing community affiliation inference in social graphs without altering topology
• Hiding community membership for overlapping community detection algorithms
• Developing deep reinforcement learning approach using proxy nodes

Innovation

• Deep reinforcement learning for community privacy
• Proxy node injection to hide membership
• Preserving graph structure while modifying edges