Improving Your Graph Neural Networks: A High-Frequency Booster

TL;DR

This paper introduces a high-frequency boosting method called CLAR to improve GNN performance on heterophilic graphs by incorporating high-pass filtering, addressing over-smoothing issues, and enhancing robustness.

Contribution

It proposes Complement Laplacian Regularization (CLAR), a novel technique that effectively integrates high-frequency information into GNNs to overcome over-smoothing and heterophily challenges.

Findings

CLAR improves GNN accuracy by up to 3.6% on benchmark datasets.

The method enhances GNN robustness to topological variations.

CLAR effectively mitigates over-smoothing in GNNs.

Abstract

Graph neural networks (GNNs) hold the promise of learning efficient representations of graph-structured data, and one of its most important applications is semi-supervised node classification. However, in this application, GNN frameworks tend to fail due to the following issues: over-smoothing and heterophily. The most popular GNNs are known to be focused on the message-passing framework, and recent research shows that these GNNs are often bounded by low-pass filters from a signal processing perspective. We thus incorporate high-frequency information into GNNs to alleviate this genetic problem. In this paper, we argue that the complement of the original graph incorporates a high-pass filter and propose Complement Laplacian Regularization (CLAR) for an efficient enhancement of high-frequency components. The experimental results demonstrate that CLAR helps GNNs tackle over-smoothing, improving the expressiveness of heterophilic graphs, which adds up to 3.6% improvement over popular baselines and ensures topological robustness.