LGCRMay 17, 2024

Rethinking Graph Backdoor Attacks: A Distribution-Preserving Perspective

arXiv:2405.10757v345 citationsh-index: 18KDD
Originality Highly original
AI Analysis

This addresses a security vulnerability in Graph Neural Networks for applications like social networks or recommendation systems, though it is incremental by focusing on making attacks less detectable.

The paper tackles the problem of existing graph backdoor attacks being easily detectable due to out-of-distribution triggers, and proposes a method to generate in-distribution triggers that bypass defenses while maintaining high attack success rates, as demonstrated on real-world datasets.

Graph Neural Networks (GNNs) have shown remarkable performance in various tasks. However, recent works reveal that GNNs are vulnerable to backdoor attacks. Generally, backdoor attack poisons the graph by attaching backdoor triggers and the target class label to a set of nodes in the training graph. A GNN trained on the poisoned graph will then be misled to predict test nodes attached with trigger to the target class. Despite their effectiveness, our empirical analysis shows that triggers generated by existing methods tend to be out-of-distribution (OOD), which significantly differ from the clean data. Hence, these injected triggers can be easily detected and pruned with widely used outlier detection methods in real-world applications. Therefore, in this paper, we study a novel problem of unnoticeable graph backdoor attacks with in-distribution (ID) triggers. To generate ID triggers, we introduce an OOD detector in conjunction with an adversarial learning strategy to generate the attributes of the triggers within distribution. To ensure a high attack success rate with ID triggers, we introduce novel modules designed to enhance trigger memorization by the victim model trained on poisoned graph. Extensive experiments on real-world datasets demonstrate the effectiveness of the proposed method in generating in distribution triggers that can by-pass various defense strategies while maintaining a high attack success rate.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes