SAIL: Self-Augmented Graph Contrastive Learning

TL;DR

SAIL introduces a self-augmented graph contrastive learning framework that enhances unsupervised node representation learning in GNNs through self-distillation, achieving competitive results across multiple benchmarks.

Contribution

The paper proposes SAIL, a novel self-augmented contrastive learning method with self-distilling modules, improving GNN performance without deep architectures.

Findings

SAIL outperforms state-of-the-art methods on benchmark datasets.

Even with a single GNN layer, SAIL achieves competitive results.

Theoretical analysis links GNN performance to feature smoothness and graph locality.

Abstract

This paper studies learning node representations with graph neural networks (GNNs) for unsupervised scenario. Specifically, we derive a theoretical analysis and provide an empirical demonstration about the non-steady performance of GNNs over different graph datasets, when the supervision signals are not appropriately defined. The performance of GNNs depends on both the node feature smoothness and the locality of graph structure. To smooth the discrepancy of node proximity measured by graph topology and node feature, we proposed SAIL - a novel \underline{S}elf-\underline{A}ugmented graph contrast\underline{i}ve \underline{L}earning framework, with two complementary self-distilling regularization modules, \emph{i.e.}, intra- and inter-graph knowledge distillation. We demonstrate the competitive performance of SAIL on a variety of graph applications. Even with a single GNN layer, SAIL has consistently competitive or even better performance on various benchmark datasets, comparing with state-of-the-art baselines.