Contrastive Learning for Non-Local Graphs with Multi-Resolution Structural Views
This work addresses the challenge of capturing structural similarity in heterophilic graphs for applications like fraudster detection and protein function prediction, representing an incremental advancement over existing contrastive methods.
The paper tackles the problem of learning node-level representations for heterophilic graphs, where existing contrastive methods fail to capture higher-order structures, and proposes a multiview contrastive learning approach that integrates diffusion filters, resulting in performance improvements such as surpassing the best baseline by 16.06% on Cornell, 3.27% on Texas, and 8.04% on Wisconsin.
Learning node-level representations of heterophilic graphs is crucial for various applications, including fraudster detection and protein function prediction. In such graphs, nodes share structural similarity identified by the equivalence of their connectivity which is implicitly encoded in the form of higher-order hierarchical information in the graphs. The contrastive methods are popular choices for learning the representation of nodes in a graph. However, existing contrastive methods struggle to capture higher-order graph structures. To address this limitation, we propose a novel multiview contrastive learning approach that integrates diffusion filters on graphs. By incorporating multiple graph views as augmentations, our method captures the structural equivalence in heterophilic graphs, enabling the discovery of hidden relationships and similarities not apparent in traditional node representations. Our approach outperforms baselines on synthetic and real structural datasets, surpassing the best baseline by $16.06\%$ on Cornell, $3.27\%$ on Texas, and $8.04\%$ on Wisconsin. Additionally, it consistently achieves superior performance on proximal tasks, demonstrating its effectiveness in uncovering structural information and improving downstream applications.