Adapting to Heterophilic Graph Data with Structure-Guided Neighbor Discovery

TL;DR

This paper introduces a novel approach called Structure-Guided GNN (SG-GNN) that enhances graph neural network performance on heterophilic data by creating and leveraging alternative structural graphs based on node similarities.

Contribution

The paper proposes a new architecture, SG-GNN, which adaptively combines original and structurally guided graphs to improve GNN performance on heterophilic datasets, supported by theoretical and empirical evidence.

Findings

SG-GNN outperforms existing methods on heterophilic benchmarks.

Creating structural graphs based on node similarity improves label homophily.

Theoretical analysis shows fewer false positive edges enhance GNN performance.

Abstract

Graph Neural Networks (GNNs) often struggle with heterophilic data, where connected nodes may have dissimilar labels, as they typically assume homophily and rely on local message passing. To address this, we propose creating alternative graph structures by linking nodes with similar structural attributes (e.g., role-based or global), thereby fostering higher label homophily on these new graphs. We theoretically prove that GNN performance can be improved by utilizing graphs with fewer false positive edges (connections between nodes of different classes) and that considering multiple graph views increases the likelihood of finding such beneficial structures. Building on these insights, we introduce Structure-Guided GNN (SG-GNN), an architecture that processes the original graph alongside the newly created structural graphs, adaptively learning to weigh their contributions. Extensive experiments on various benchmark datasets, particularly those with heterophilic characteristics, demonstrate that our SG-GNN achieves state-of-the-art or highly competitive performance, highlighting the efficacy of exploiting structural information to guide GNNs.