LOHA: Direct Graph Spectral Contrastive Learning Between Low-pass and High-pass Views
This work addresses a gap in spectral GNNs for graphs with varying homophily levels, offering a novel method that is incremental but shows strong gains.
The paper tackles the problem of suboptimal performance in spectral graph neural networks by proposing LOHA, a self-supervised contrastive framework that harmonizes low-pass and high-pass views, achieving an average 2.8% improvement over runner-up models on 9 datasets and surpassing fully-supervised models on several.
Spectral Graph Neural Networks effectively handle graphs with different homophily levels, with low-pass filter mining feature smoothness and high-pass filter capturing differences. When these distinct filters could naturally form two opposite views for self-supervised learning, the commonalities between the counterparts for the same node remain unexplored, leading to suboptimal performance. In this paper, a simple yet effective self-supervised contrastive framework, LOHA, is proposed to address this gap. LOHA optimally leverages low-pass and high-pass views by embracing "harmony in diversity". Rather than solely maximizing the difference between these distinct views, which may lead to feature separation, LOHA harmonizes the diversity by treating the propagation of graph signals from both views as a composite feature. Specifically, a novel high-dimensional feature named spectral signal trend is proposed to serve as the basis for the composite feature, which remains relatively unaffected by changing filters and focuses solely on original feature differences. LOHA achieves an average performance improvement of 2.8% over runner-up models on 9 real-world datasets with varying homophily levels. Notably, LOHA even surpasses fully-supervised models on several datasets, which underscores the potential of LOHA in advancing the efficacy of spectral GNNs for diverse graph structures.