Attacking and Securing Community Detection: A Game-Theoretic Framework

2025-12-12

AI Summary

This paper addresses the dual challenges of model failure under adversarial graph attacks and target-node privacy leakage in community detection. We propose CD-GAME, the first game-theoretic framework for joint robustness and privacy modeling in this setting. CD-GAME formalizes a multi-round dynamic game between an attacker—aiming to covertly hide target nodes—and a Rayleigh-quotient-driven defender, converging to a Nash equilibrium. Our contributions include: (i) the first integration of game theory into community detection for simultaneous robustness and privacy guarantees; (ii) a stealthy yet highly effective dynamic attack strategy and an adaptive defense mechanism; and (iii) theoretical and empirical insights distinguishing single-step attacks/defenses from equilibrium strategies. Extensive experiments on real-world and synthetic graphs demonstrate that CD-GAME significantly outperforms baselines in attack invisibility, defense robustness (substantial NMI/F1 gains), and resilience against countermeasures—establishing a new paradigm for trustworthy community analysis with graph neural networks.

Technology Category

• Game Theory and Economic Paradigms: Adversarial Learning
• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Security and Privacy: Large-scale security measurements

Abstract

It has been demonstrated that adversarial graphs, i.e., graphs with imperceptible perturbations, can cause deep graph models to fail on classification tasks. In this work, we extend the concept of adversarial graphs to the community detection problem, which is more challenging. We propose novel attack and defense techniques for community detection problem, with the objective of hiding targeted individuals from detection models and enhancing the robustness of community detection models, respectively. These techniques have many applications in real-world scenarios, for example, protecting personal privacy in social networks and understanding camouflage patterns in transaction networks. To simulate interactive attack and defense behaviors, we further propose a game-theoretic framework, called CD-GAME. One player is a graph attacker, while the other player is a Rayleigh Quotient defender. The CD-GAME models the mutual influence and feedback mechanisms between the attacker and the defender, revealing the dynamic evolutionary process of the game. Both players dynamically update their strategies until they reach the Nash equilibrium. Extensive experiments demonstrate the effectiveness of our proposed attack and defense methods, and both outperform existing baselines by a significant margin. Furthermore, CD-GAME provides valuable insights for understanding interactive attack and defense scenarios in community detection problems. We found that in traditional single-step attack or defense, attacker tends to employ strategies that are most effective, but are easily detected and countered by defender. When the interactive game reaches a Nash equilibrium, attacker adopts more imperceptible strategies that can still achieve satisfactory attack effectiveness even after defense.

Problem

• Extends adversarial graphs to community detection challenges
• Proposes attack and defense techniques for hiding individuals and enhancing robustness
• Introduces a game-theoretic framework modeling interactive attack-defense dynamics

Innovation

• Game-theoretic framework models attack-defense interactions
• Novel techniques for hiding individuals and enhancing robustness
• Dynamic strategy updates until Nash equilibrium reached