Node Injection Attack Based on Label Propagation Against Graph Neural Network

TL;DR

This paper introduces LPGIA, a novel label-propagation-based graph injection attack that effectively compromises GNNs by injecting malicious nodes, outperforming existing methods across multiple datasets.

Contribution

The paper proposes LPGIA, a new attack method targeting the label propagation process in GNNs, filling a gap in existing graph injection attack research.

Findings

LPGIA outperforms previous attack methods in various datasets.

LPGIA demonstrates high transferability across different GNN models.

The attack effectively exploits label propagation to disrupt node classification.

Abstract

Graph Neural Network (GNN) has achieved remarkable success in various graph learning tasks, such as node classification, link prediction and graph classification. The key to the success of GNN lies in its effective structure information representation through neighboring aggregation. However, the attacker can easily perturb the aggregation process through injecting fake nodes, which reveals that GNN is vulnerable to the graph injection attack. Existing graph injection attack methods primarily focus on damaging the classical feature aggregation process while overlooking the neighborhood aggregation process via label propagation. To bridge this gap, we propose the label-propagation-based global injection attack (LPGIA) which conducts the graph injection attack on the node classification task. Specifically, we analyze the aggregation process from the perspective of label propagation and transform the graph injection attack problem into a global injection label specificity attack problem. To solve this problem, LPGIA utilizes a label propagation-based strategy to optimize the combinations of the nodes connected to the injected node. Then, LPGIA leverages the feature mapping to generate malicious features for injected nodes. In extensive experiments against representative GNNs, LPGIA outperforms the previous best-performing injection attack method in various datasets, demonstrating its superiority and transferability.