Are defenses for graph neural networks robust?

F Mujkanovic, S Geisler… - Advances in Neural …, 2022 - proceedings.neurips.cc
A cursory reading of the literature suggests that we have made a lot of progress in designing
effective adversarial defenses for Graph Neural Networks (GNNs). Yet, the standard …

Dos-gnn: Dual-feature aggregations with over-sampling for class-imbalanced fraud detection on graphs

S **g, L Chen, Q Li, D Wu - 2024 International Joint …, 2024 - ieeexplore.ieee.org
As fraudulent activities have shot up manifolds, fraud detection has emerged as a pivotal
process in different fields (eg, e-commerce, online reviews, and social networks). Since …

Distilling knowledge on text graph for social media attribute inference

Q Li, X Li, L Chen, D Wu - Proceedings of the 45th International ACM …, 2022 - dl.acm.org
The popularization of social media generates a large amount of user-oriented data, where
text data especially attracts researchers and speculators to infer user attributes (eg, age …

Hover: Homophilic oversampling via edge removal for class-imbalanced bot detection on graphs

B Ashmore, L Chen - Proceedings of the 32nd ACM International …, 2023 - dl.acm.org
As malicious bots reside in a network to disrupt network stability, graph neural networks
(GNNs) have emerged as one of the most popular bot detection methods. However, in most …

Adversarial diffusion attacks on graph-based traffic prediction models

L Zhu, K Feng, Z Pu, W Ma - IEEE Internet of Things Journal, 2023 - ieeexplore.ieee.org
Real-time traffic prediction models play a pivotal role in smart mobility systems and have
been widely used in route guidance, emerging mobility services, and advanced traffic …

Adversary for social good: Leveraging attribute-obfuscating attack to protect user privacy on social networks

X Li, L Chen, D Wu - International Conference on Security and Privacy in …, 2022 - Springer
As social networks become indispensable for people's daily lives, inference attacks pose
significant threat to users' privacy where attackers can infiltrate users' information and infer …

Pseudo-Labeling with Graph Active Learning for Few-shot Node Classification

Q Li, L Chen, S **g, D Wu - 2023 IEEE International …, 2023 - ieeexplore.ieee.org
Graphs have emerged as one of the most important and powerful data structures to perform
content analysis in many fields. In this line of work, node classification is a classic task …

Hierarchical graph neural network for patient treatment preference prediction with external knowledge

Q Li, L Chen, Y Cai, D Wu - … Asia Conference on Knowledge Discovery and …, 2023 - Springer
The healthcare industry has a wealth of data that can be used by researchers and medical
professionals to infer a patient's condition and intention to receive treatment using machine …

Towards defense against adversarial attacks on graph neural networks via calibrated co-training

XG Wu, HJ Wu, X Zhou, X Zhao, K Lu - Journal of Computer Science and …, 2022 - Springer
Graph neural networks (GNNs) have achieved significant success in graph representation
learning. Nevertheless, the recent work indicates that current GNNs are vulnerable to …

A comparative study on robust graph neural networks to structural noises

Z Zhang, Y Pei - arxiv preprint arxiv:2112.06070, 2021 - arxiv.org
Graph neural networks (GNNs) learn node representations by passing and aggregating
messages between neighboring nodes. GNNs have been applied successfully in several …