Adversarial Robustness of Link Sign Prediction in Signed Graphs

TL;DR

This paper investigates the vulnerabilities of signed graph neural networks to adversarial attacks, introduces a novel attack method, and proposes a robust contrastive learning framework to enhance model resilience in social network analysis.

Contribution

It introduces balance-attack, a new adversarial strategy, and BA-SGCL, a contrastive learning framework that improves robustness against such attacks in signed graph neural networks.

Findings

Balance-attack effectively compromises graph balance degree.

BA-SGCL significantly improves robustness across multiple SGNN architectures.

Experimental results demonstrate enhanced security and reliability in signed graph analysis.

Abstract

Signed graphs serve as fundamental data structures for representing positive and negative relationships in social networks, with signed graph neural networks (SGNNs) emerging as the primary tool for their analysis. Our investigation reveals that balance theory, while essential for modeling signed relationships in SGNNs, inadvertently introduces exploitable vulnerabilities to black-box attacks. To showcase this, we propose balance-attack, a novel adversarial strategy specifically designed to compromise graph balance degree, and develop an efficient heuristic algorithm to solve the associated NP-hard optimization problem. While existing approaches attempt to restore attacked graphs through balance learning techniques, they face a critical challenge we term "Irreversibility of Balance-related Information," as restored edges fail to align with original attack targets. To address this limitation, we introduce Balance Augmented-Signed Graph Contrastive Learning (BA-SGCL), an innovative framework that combines contrastive learning with balance augmentation techniques to achieve robust graph representations. By maintaining high balance degree in the latent space, BA-SGCL not only effectively circumvents the irreversibility challenge but also significantly enhances model resilience. Extensive experiments across multiple SGNN architectures and real-world datasets demonstrate both the effectiveness of our proposed balance-attack and the superior robustness of BA-SGCL, advancing the security and reliability of signed graph analysis in social networks. Datasets and codes of the proposed framework are at the github repository https://anonymous.4open.science/r/BA-SGCL-submit-DF41/.