Adversarial Defense Framework for Graph Neural Network

Abstract

Graph neural network (GNN), as a powerful representation learning model on graph data, attracts much attention across various disciplines. However, recent studies show that GNN is vulnerable to adversarial attacks. How to make GNN more robust? What are the key vulnerabilities in GNN? How to address the vulnerabilities and defense GNN against the adversarial attacks? In this paper, we propose DefNet, an effective adversarial defense framework for GNNs. In particular, we first investigate the latent vulnerabilities in every layer of GNNs and propose corresponding strategies including dual-stage aggregation and bottleneck perceptron. Then, to cope with the scarcity of training data, we propose an adversarial contrastive learning method to train the GNN in a conditional GAN manner by leveraging the high-level graph representation. Extensive experiments on three public datasets demonstrate the effectiveness of DefNet in improving the robustness of popular GNN variants, such as Graph Convolutional Network and GraphSAGE, under various types of adversarial attacks.

References in corpus (11)

• Semi-Supervised Classification with Graph Convolutional Networks
• Inductive Representation Learning on Large Graphs
• Towards Deep Learning Models Resistant to Adversarial Attacks
• Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
• Adversarial Attacks on Neural Networks for Graph Data
• Gated Graph Sequence Neural Networks
• How Powerful are Graph Neural Networks?
• Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models
• Adversarial Examples for Evaluating Reading Comprehension Systems
• Adversarial Attack on Graph Structured Data
• Cascade Adversarial Machine Learning Regularized with a Unified Embedding

Cited by in corpus (10)

• Adversarial Attack and Defense on Graph Data: A Survey
• GNNGuard: Defending Graph Neural Networks against Adversarial Attacks
• When Does Self-Supervision Help Graph Convolutional Networks?
• A Survey of Adversarial Learning on Graphs
• GraphDefense: Towards Robust Graph Convolutional Networks
• Understanding and Improvement of Adversarial Training for Network Embedding from an Optimization Perspective
• Learning Robust Representation through Graph Adversarial Contrastive Learning
• Heterogeneous Graph Matching Networks
• AN-GCN: An Anonymous Graph Convolutional Network Defense Against Edge-Perturbing Attack
• Spatio-Temporal Sparsification for General Robust Graph Convolution Networks