Similarity Preserving Adversarial Graph Contrastive Learning

TL;DR

This paper introduces SP-AGCL, a novel adversarial graph contrastive learning framework that maintains node feature similarity while enhancing robustness against adversarial attacks, noise, and heterophily.

Contribution

It proposes a new contrastive learning method that preserves node similarity and improves adversarial robustness in graph neural networks.

Findings

SP-AGCL outperforms existing methods on downstream tasks.

It is effective against adversarial attacks, noisy labels, and heterophilous neighbors.

The framework maintains node feature similarity while enhancing robustness.

Abstract

Recent works demonstrate that GNN models are vulnerable to adversarial attacks, which refer to imperceptible perturbation on the graph structure and node features. Among various GNN models, graph contrastive learning (GCL) based methods specifically suffer from adversarial attacks due to their inherent design that highly depends on the self-supervision signals derived from the original graph, which however already contains noise when the graph is attacked. To achieve adversarial robustness against such attacks, existing methods adopt adversarial training (AT) to the GCL framework, which considers the attacked graph as an augmentation under the GCL framework. However, we find that existing adversarially trained GCL methods achieve robustness at the expense of not being able to preserve the node feature similarity. In this paper, we propose a similarity-preserving adversarial graph contrastive learning (SP-AGCL) framework that contrasts the clean graph with two auxiliary views of different properties (i.e., the node similarity-preserving view and the adversarial view). Extensive experiments demonstrate that SP-AGCL achieves a competitive performance on several downstream tasks, and shows its effectiveness in various scenarios, e.g., a network with adversarial attacks, noisy labels, and heterophilous neighbors. Our code is available at https://github.com/yeonjun-in/torch-SP-AGCL.