Bayesian Node Classification for Noisy Graphs

Hakim Hafidi, Mounir Ghogho, Philippe Ciblat, Ananthram Swami

Abstract: Graph neural networks (GNN) have been recognized as powerful tools for learning representations in graph structured data. The key idea is to propagate and aggregate information along edges of the given graph. However, little work has been done to analyze the effect of noise on their performance. By conducting a number of simulations, we show that GNN are very sensitive to graph noise. We propose a graph-assisted Bayesian node classifier which takes into account the degree of impurity of the graph, and show that it consistently outperforms GNN based classifiers on benchmark datasets, particularly when the degree of impurity is moderate to high.