Dual-Kernel Graph Community Contrastive Learning

TL;DR

This paper introduces an efficient graph contrastive learning framework that leverages community structures and kernelized loss to improve scalability and performance of GNNs on large graphs.

Contribution

It proposes a novel kernelized contrastive loss with linear complexity and integrates knowledge distillation into GCL for faster inference and better scalability.

Findings

Outperforms state-of-the-art GCL methods on multiple datasets.

Achieves linear complexity in contrastive loss computation.

Demonstrates improved scalability and effectiveness on large-scale graphs.

Abstract

Graph Contrastive Learning (GCL) has emerged as a powerful paradigm for training Graph Neural Networks (GNNs) in the absence of task-specific labels. However, its scalability on large-scale graphs is hindered by the intensive message passing mechanism of GNN and the quadratic computational complexity of contrastive loss over positive and negative node pairs. To address these issues, we propose an efficient GCL framework that transforms the input graph into a compact network of interconnected node sets while preserving structural information across communities. We firstly introduce a kernelized graph community contrastive loss with linear complexity, enabling effective information transfer among node sets to capture hierarchical structural information of the graph. We then incorporate a knowledge distillation technique into the decoupled GNN architecture to accelerate inference while maintaining strong generalization performance. Extensive experiments on sixteen real-world datasets of varying scales demonstrate that our method outperforms state-of-the-art GCL baselines in both effectiveness and scalability.