Adversarial Attacks on Node Embeddings via Graph Poisoning

Abstract

The goal of network representation learning is to learn low-dimensional node embeddings that capture the graph structure and are useful for solving downstream tasks. However, despite the proliferation of such methods, there is currently no study of their robustness to adversarial attacks. We provide the first adversarial vulnerability analysis on the widely used family of methods based on random walks. We derive efficient adversarial perturbations that poison the network structure and have a negative effect on both the quality of the embeddings and the downstream tasks. We further show that our attacks are transferable since they generalize to many models and are successful even when the attacker is restricted.

arXiv commentICML 2019, PMLR 97:695-704

Cited by in corpus (41)

• Adversarial Attack and Defense on Graph Data: A Survey
• Transferring Robustness for Graph Neural Network Against Poisoning Attacks
• GNNGuard: Defending Graph Neural Networks against Adversarial Attacks
• Topology Attack and Defense for Graph Neural Networks: An Optimization Perspective
• Single Node Injection Attack against Graph Neural Networks
• Node-Level Membership Inference Attacks Against Graph Neural Networks
• Graph Structure Learning for Robust Graph Neural Networks
• Stealing Links from Graph Neural Networks
• Adversarial Attack Framework on Graph Embedding Models with Limited Knowledge
• Reliable Graph Neural Networks via Robust Aggregation
• Link Prediction Adversarial Attack
• Can Adversarial Network Attack be Defended?
• Residual Network and Embedding Usage: New Tricks of Node Classification with Graph Convolutional Networks
• Towards More Practical Adversarial Attacks on Graph Neural Networks
• Surrogate Representation Learning with Isometric Mapping for Gray-box Graph Adversarial Attacks
• Attacking Graph-based Classification via Manipulating the Graph Structure
• Backdoor Attacks to Graph Neural Networks
• Towards Security Threats of Deep Learning Systems: A Survey
• Learning to Deceive Knowledge Graph Augmented Models via Targeted Perturbation
• Efficient Robustness Certificates for Discrete Data: Sparsity-Aware Randomized Smoothing for Graphs, Images and More
• Information Obfuscation of Graph Neural Networks
• Adversarial Attacks on Graph Classification via Bayesian Optimisation
• Certified Robustness of Community Detection against Adversarial Structural Perturbation via Randomized Smoothing
• Multiscale Evolutionary Perturbation Attack on Community Detection
• A Restricted Black-box Adversarial Framework Towards Attacking Graph Embedding Models
• Adversarial Attack on Network Embeddings via Supervised Network Poisoning
• Understanding Structural Vulnerability in Graph Convolutional Networks
• A Targeted Universal Attack on Graph Convolutional Network
• Certified Robustness of Graph Neural Networks against Adversarial Structural Perturbation
• Adversarial Attack on Large Scale Graph
• Stacked Graph Filter
• ANDRUSPEX : Leveraging Graph Representation Learning to Predict Harmful App Installations on Mobile Devices
• Topological Effects on Attacks Against Vertex Classification
• CollaborER: A Self-supervised Entity Resolution Framework Using Multi-features Collaboration
• Optimal Edge Weight Perturbations to Attack Shortest Paths
• CoG: a Two-View Co-training Framework for Defending Adversarial Attacks on Graph
• Network representation learning: A macro and micro view
• AN-GCN: An Anonymous Graph Convolutional Network Defense Against Edge-Perturbing Attack
• Network Representation Learning: From Traditional Feature Learning to Deep Learning
• GA Based Q-Attack on Community Detection
• A Le Cam Type Bound for Adversarial Learning and Applications