Revisiting Edge Perturbation for Graph Neural Network in Graph Data Augmentation and Attack

arXiv:2403.07943v1 Announce Type: new
Abstract: Edge perturbation is a basic method to modify graph structures. It can be categorized into two veins based on their effects on the performance of graph neural networks (GNNs), i.e., graph data augmentation and attack. Surprisingly, both veins of edge perturbation methods employ the same operations, yet yield opposite effects on GNNs' accuracy. A distinct boundary between these methods in using edge perturbation has never been clearly defined. Consequently, inappropriate perturbations may lead to undesirable …

abstract arxiv augmentation basic cs.cr cs.lg data edge effects gnns graph graph data graph neural network graph neural networks network networks neural network neural networks performance type