Time: 4:30 pm
Venue: YEUNG-B5311, City University of Hong Kong
Zoom Meeting ID: 859 2865 1236
Password: 123456
Speaker: Mr Junyuan Fang, City University of Hong Kong
Mitigating the Structural Bias in Graph Adversarial Defenses
Mr Junyuan Fang, City University of Hong Kong
Abstract: In recent years, graph neural networks (GNNs) have shown great potential in addressing various graph structure-related downstream tasks. However, recent studies have found that current GNNs are susceptible to malicious adversarial attacks. Given the inevitable presence of adversarial attacks in the real world, a variety of defense methods have been proposed to counter these attacks and enhance the robustness of GNNs. Despite the commendable performance of these defense methods, we have observed that they tend to exhibit a structural bias in terms of their defense capability on nodes with low degree (i.e., tail nodes), which is similar to the structural bias of traditional GNNs on nodes with low degree in the clean graph. Therefore, in this talk, we propose a defense strategy by including hetero-homo augmented graph construction, GNN augmented graph construction, and multi-view node-wise attention modules to mitigate the structural bias of GNNs against adversarial attacks. Notably, the hetero-homo augmented graph consists of removing heterophilic links (i.e., links connecting nodes with dissimilar features) globally and adding homophilic links (i.e., links connecting nodes with similar features) for nodes with low degree. To further enhance the defense capability, an attention mechanism is adopted to adaptively combine the representations from the above two kinds of graph views. We conduct extensive experiments to demonstrate the defense and debiasing effect of the proposed strategy on benchmark datasets.
Speaker’s Bio:Junyuan Fang received the B.Eng. degree in software engineering from the Guangdong University of Technology in 2018, the M.Eng. degree in software engineering from Sun Yat-sen University in 2020, and the Ph.D. degree in electrical engineering from the City University of Hong Kong in 2024. His current research interests include robustness analysis and optimization, applications of network science, and graph mining.
WEBINAR WEBSITE:
