HTG-GCL: Leveraging Hierarchical Topological Granularity from Cellular Complexes for Graph Contrastive Learning

TL;DR

HTG-GCL introduces a hierarchical approach to graph contrastive learning by leveraging multi-scale topological views from cellular complexes, improving the capture of meaningful graph representations across different tasks.

Contribution

The paper proposes HTG-GCL, a novel framework that generates multi-scale topological views and applies a multi-granularity contrastive learning mechanism with uncertainty-based weighting.

Findings

Outperforms existing GCL methods on multiple benchmarks.

Effectively captures hierarchical topological information.

Demonstrates robustness across various downstream tasks.

Abstract

Graph contrastive learning (GCL) aims to learn discriminative semantic invariance by contrasting different views of the same graph that share critical topological patterns. However, existing GCL approaches with structural augmentations often struggle to identify task-relevant topological structures, let alone adapt to the varying coarse-to-fine topological granularities required across different downstream tasks. To remedy this issue, we introduce Hierarchical Topological Granularity Graph Contrastive Learning (HTG-GCL), a novel framework that leverages transformations of the same graph to generate multi-scale ring-based cellular complexes, embodying the concept of topological granularity, thereby generating diverse topological views. Recognizing that a certain granularity may contain misleading semantics, we propose a multi-granularity decoupled contrast and apply a granularity-specific weighting mechanism based on uncertainty estimation. Comprehensive experiments on various benchmarks demonstrate the effectiveness of HTG-GCL, highlighting its superior performance in capturing meaningful graph representations through hierarchical topological information.