Graph Poisoning Attack

Graph poisoning attacks target the vulnerabilities of graph neural networks (GNNs) by subtly altering the graph structure—adding or removing edges—to degrade their performance on downstream tasks like node classification or link prediction. Current research focuses on developing more effective attack strategies, including those that address biases in existing methods and optimize resource allocation for maximum impact, often employing gradient-based approaches and contrastive learning techniques. These attacks highlight the critical need for robust GNNs, impacting the reliability of applications across various domains that rely on graph-structured data, such as social network analysis and fraud detection, and driving the development of effective defense mechanisms.

Papers

July 27, 2024

Debiased Graph Poisoning Attack via Contrastive Surrogate Objective
Kanghoon Yoon, Yeonjun In, Namkyeong Lee, Kibum Kim, Chanyoung Park
Adversarial Attack Gradient Based Attack Meta Gradient De Bias Surrogate Objective Graph Poisoning Attack

September 19, 2023

Crypto'Graph: Leveraging Privacy-Preserving Distributed Link Prediction for Robust Graph Learning
Sofiane Azogagh, Zelma Aubin Birba, Sébastien Gambs, Marc-Olivier Killijian
Privacy Preserving Cryptographic Primitive Graph Poisoning Attack Key Graph

August 15, 2023

Simple and Efficient Partial Graph Adversarial Attack: A New Perspective
Guanghui Zhu, Mengyu Chen, Chunfeng Yuan, Yihua Huang
New Perspective Malicious Node Graph Adversarial Attack Graph Poisoning Attack

August 2, 2023

VertexSerum: Poisoning Graph Neural Networks for Link Inference
Ruyi Ding, Shijin Duan, Xiaolin Xu, Yunsi Fei
Graph Data Goal Vertex Attribute Inference Attack Graph Poisoning Attack

March 29, 2023

Towards Reasonable Budget Allocation in Untargeted Graph Structure Attacks via Gradient Debias
Zihan Liu, Yun Luo, Lirong Wu, Zicheng Liu, Stan Z. Li
Cross Entropy Loss Cross Entropy Untargeted Attack Budget Allocation Graph Poisoning Attack

January 20, 2022

Unsupervised Graph Poisoning Attack via Contrastive Loss Back-propagation
Sixiao Zhang, Hongxu Chen, Xiangguo Sun, Yicong Li, Guandong Xu
Contrastive Learning Adversarial Attack Graph Representation Learning Supervised Contrastive Loss Gradient Based Adversarial Graph Poisoning Attack

January 17, 2022

Neighboring Backdoor Attacks on Graph Convolutional Network
Liang Chen, Qibiao Peng, Jintang Li, Yang Liu, Jiawei Chen, Yong Li, Zibin Zheng
Backdoor Attack Graph Convolutional Network Graph Backdoor Attack Graph Poisoning Attack