Robust Node Classification on Graphs: Jointly from Bayesian Label Transition and Topology-based Label Propagation
This work addresses robustness issues in graph neural networks for node classification, which is important for applications in real-world scenarios, but it is incremental as it builds on existing label transition methods.
The paper tackles the problem of GNN-based node classification performance deteriorating under topological perturbations by proposing LInDT, a label inference model that integrates Bayesian label transition and topology-based label propagation, which demonstrates superiority in experiments on five datasets under three perturbation scenarios.
Node classification using Graph Neural Networks (GNNs) has been widely applied in various real-world scenarios. However, in recent years, compelling evidence emerges that the performance of GNN-based node classification may deteriorate substantially by topological perturbation, such as random connections or adversarial attacks. Various solutions, such as topological denoising methods and mechanism design methods, have been proposed to develop robust GNN-based node classifiers but none of these works can fully address the problems related to topological perturbations. Recently, the Bayesian label transition model is proposed to tackle this issue but its slow convergence may lead to inferior performance. In this work, we propose a new label inference model, namely LInDT, which integrates both Bayesian label transition and topology-based label propagation for improving the robustness of GNNs against topological perturbations. LInDT is superior to existing label transition methods as it improves the label prediction of uncertain nodes by utilizing neighborhood-based label propagation leading to better convergence of label inference. Besides, LIndT adopts asymmetric Dirichlet distribution as a prior, which also helps it to improve label inference. Extensive experiments on five graph datasets demonstrate the superiority of LInDT for GNN-based node classification under three scenarios of topological perturbations.