Contrastive Multi-View Representation Learning on Graphs

TL;DR

This paper presents a self-supervised contrastive learning method for graph representations, achieving state-of-the-art results by contrasting structural views like first-order neighbors and graph diffusion, outperforming previous methods.

Contribution

The paper introduces a novel contrastive learning approach for graphs that identifies the most effective views, achieving new state-of-the-art results on multiple benchmarks.

Findings

Achieves 86.8% accuracy on Cora node classification

Achieves 84.5% accuracy on Reddit-Binary graph classification

Contrasting first-order neighbors and diffusion yields best performance

Abstract

We introduce a self-supervised approach for learning node and graph level representations by contrasting structural views of graphs. We show that unlike visual representation learning, increasing the number of views to more than two or contrasting multi-scale encodings do not improve performance, and the best performance is achieved by contrasting encodings from first-order neighbors and a graph diffusion. We achieve new state-of-the-art results in self-supervised learning on 8 out of 8 node and graph classification benchmarks under the linear evaluation protocol. For example, on Cora (node) and Reddit-Binary (graph) classification benchmarks, we achieve 86.8% and 84.5% accuracy, which are 5.5% and 2.4% relative improvements over previous state-of-the-art. When compared to supervised baselines, our approach outperforms them in 4 out of 8 benchmarks. Source code is released at: https://github.com/kavehhassani/mvgrl