Partition-wise Graph Filtering: A Unified Perspective Through the Lens of Graph Coarsening

TL;DR

This paper introduces a unified graph filtering framework called Coarsening-guided Partition-wise Filtering (CPF), which combines graph coarsening and feature clustering to improve adaptability and performance of graph neural networks on diverse graph types.

Contribution

It proposes a novel CPF method that performs filtering on node partitions, unifying graph-wise and node-wise filtering paradigms with theoretical analysis and empirical validation.

Findings

CPF outperforms existing filtering paradigms in node classification tasks.

The method effectively detects anomalies in real-world graphs.

Theoretical analysis shows CPF's superiority over traditional filtering approaches.

Abstract

Filtering-based graph neural networks (GNNs) constitute a distinct class of GNNs that employ graph filters to handle graph-structured data, achieving notable success in various graph-related tasks. Conventional methods adopt a graph-wise filtering paradigm, imposing a uniform filter across all nodes, yet recent findings suggest that this rigid paradigm struggles with heterophilic graphs. To overcome this, recent works have introduced node-wise filtering, which assigns distinct filters to individual nodes, offering enhanced adaptability. However, a fundamental gap remains: a comprehensive framework unifying these two strategies is still absent, limiting theoretical insights into the filtering paradigms. Moreover, through the lens of Contextual Stochastic Block Model, we reveal that a synthesis of graph-wise and node-wise filtering provides a sufficient solution for classification on graphs exhibiting both homophily and heterophily, suggesting the risk of excessive parameterization and potential overfitting with node-wise filtering. To address the limitations, this paper introduces Coarsening-guided Partition-wise Filtering (CPF). CPF innovates by performing filtering on node partitions. The method begins with structure-aware partition-wise filtering, which filters node partitions obtained via graph coarsening algorithms, and then performs feature-aware partition-wise filtering, refining node embeddings via filtering on clusters produced by $k$-means clustering over features. In-depth analysis is conducted for each phase of CPF, showing its superiority over other paradigms. Finally, benchmark node classification experiments, along with a real-world graph anomaly detection application, validate CPF's efficacy and practical utility.