arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.01263cs.CRcs.LG
收藏

图神经网络的不可察觉后门攻击

Unnoticeable Backdoor Attacks on Graph Neural Networks

The Pennsylvania State University(宾夕法尼亚州立大学)

Enyan Dai, Minhua Lin, Xiang Zhang, Suhang Wang

摘要:

图神经网络(GNNs)在节点分类和图分类等各种任务中取得了可喜的成果。最近的研究发现,GNN容易受到对抗攻击的攻击。然而,有效的图上的后门攻击仍然是一个悬而未决的问题。特别地,后门攻击通过将触发器和目标类标签附加到训练图中的一组节点来毒害图。然后,在中毒图上训练的后门GNN将被误导,一旦附加了触发器,就将测试节点预测为目标类。虽然有一些初步的努力,图后门攻击,我们的实证分析表明,他们可能需要一个大的攻击预算有效的后门攻击和注入的触发器可以很容易地检测和修剪。因此,在本文中,我们研究了一个新的问题,不可察觉的图后门攻击有限的攻击预算。为了充分利用攻击预算,我们建议在中毒阶段故意选择注入触发器和目标类标签的节点。采用自适应触发发生器来获取不易被注意到的有效触发。针对各种防御策略在真实数据集上进行的大量实验表明,我们提出的方法在进行有效的不明显的后门攻击方面是有效的。

英文摘要:

Graph Neural Networks (GNNs) have achieved promising results in various tasks such as node classification and graph classification. Recent studies find that GNNs are vulnerable to adversarial attacks. However, effective backdoor attacks on graphs are still an open problem. In particular, backdoor attack poisons the graph by attaching triggers and the target class label to a set of nodes in the training graph. The backdoored GNNs trained on the poisoned graph will then be misled to predict test nodes to target class once attached with triggers. Though there are some initial efforts in graph backdoor attacks, our empirical analysis shows that they may require a large attack budget for effective backdoor attacks and the injected triggers can be easily detected and pruned. Therefore, in this paper, we study a novel problem of unnoticeable graph backdoor attacks with limited attack budget. To fully utilize the attack budget, we propose to deliberately select the nodes to inject triggers and target class labels in the poisoning phase. An adaptive trigger generator is deployed to obtain effective triggers that are difficult to be noticed. Extensive experiments on real-world datasets against various defense strategies demonstrate the effectiveness of our proposed method in conducting effective unnoticeable backdoor attacks.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑