Heterophilous Distribution Propagation for Graph Neural Networks

TL;DR

This paper introduces heterophilous distribution propagation (HDP), a novel GNN method that adaptively separates and propagates homophilous and heterophilous neighborhood information, improving performance on heterophilous graphs.

Contribution

HDP is the first to adaptively partition neighbors and learn heterophilous distributions with contrastive learning, addressing key challenges in heterophilous GNNs.

Findings

HDP outperforms baselines on heterophilous datasets.

HDP effectively separates homophilous and heterophilous neighborhood information.

Experimental results validate the superiority of HDP across diverse benchmarks.

Abstract

Graph Neural Networks (GNNs) have achieved remarkable success in various graph mining tasks by aggregating information from neighborhoods for representation learning. The success relies on the homophily assumption that nearby nodes exhibit similar behaviors, while it may be violated in many real-world graphs. Recently, heterophilous graph neural networks (HeterGNNs) have attracted increasing attention by modifying the neural message passing schema for heterophilous neighborhoods. However, they suffer from insufficient neighborhood partition and heterophily modeling, both of which are critical but challenging to break through. To tackle these challenges, in this paper, we propose heterophilous distribution propagation (HDP) for graph neural networks. Instead of aggregating information from all neighborhoods, HDP adaptively separates the neighbors into homophilous and heterphilous parts based on the pseudo assignments during training. The heterophilous neighborhood distribution is learned with orthogonality-oriented constraint via a trusted prototype contrastive learning paradigm. Both the homophilous and heterophilous patterns are propagated with a novel semantic-aware message passing mechanism. We conduct extensive experiments on 9 benchmark datasets with different levels of homophily. Experimental results show that our method outperforms representative baselines on heterophilous datasets.