Black-box Node Injection Attack for Graph Neural Networks

TL;DR

This paper introduces GA2C, a black-box node injection attack framework for GNNs that uses reinforcement learning to generate and insert nodes, effectively evading GNN models without needing access to their internal structure.

Contribution

The paper presents a novel black-box node injection attack method for GNNs using reinforcement learning, which is more practical and effective than previous structural perturbation approaches.

Findings

GA2C outperforms existing attack methods on benchmark datasets.

The approach effectively evades GNNs with limited knowledge of the model.

Injected nodes are realistic and seamlessly integrated into the original graph.

Abstract

Graph Neural Networks (GNNs) have drawn significant attentions over the years and been broadly applied to vital fields that require high security standard such as product recommendation and traffic forecasting. Under such scenarios, exploiting GNN's vulnerabilities and further downgrade its classification performance become highly incentive for adversaries. Previous attackers mainly focus on structural perturbations of existing graphs. Although they deliver promising results, the actual implementation needs capability of manipulating the graph connectivity, which is impractical in some circumstances. In this work, we study the possibility of injecting nodes to evade the victim GNN model, and unlike previous related works with white-box setting, we significantly restrict the amount of accessible knowledge and explore the black-box setting. Specifically, we model the node injection attack as a Markov decision process and propose GA2C, a graph reinforcement learning framework in the fashion of advantage actor critic, to generate realistic features for injected nodes and seamlessly merge them into the original graph following the same topology characteristics. Through our extensive experiments on multiple acknowledged benchmark datasets, we demonstrate the superior performance of our proposed GA2C over existing state-of-the-art methods. The data and source code are publicly accessible at: https://github.com/jumxglhf/GA2C.