Graph Contrastive Learning with Implicit Augmentations

TL;DR

This paper introduces iGCL, a novel graph contrastive learning method that uses implicit augmentations learned from a Variational Graph Auto-Encoder, avoiding manual tuning and preserving graph semantics, leading to state-of-the-art results.

Contribution

The paper proposes a new implicit augmentation technique for graph contrastive learning using latent space representations, eliminating manual augmentation design.

Findings

Achieves state-of-the-art performance on graph and node-level tasks.

Demonstrates effectiveness of implicit augmentations over explicit perturbations.

Validates modules of iGCL through ablation studies.

Abstract

Existing graph contrastive learning methods rely on augmentation techniques based on random perturbations (e.g., randomly adding or dropping edges and nodes). Nevertheless, altering certain edges or nodes can unexpectedly change the graph characteristics, and choosing the optimal perturbing ratio for each dataset requires onerous manual tuning. In this paper, we introduce Implicit Graph Contrastive Learning (iGCL), which utilizes augmentations in the latent space learned from a Variational Graph Auto-Encoder by reconstructing graph topological structure. Importantly, instead of explicitly sampling augmentations from latent distributions, we further propose an upper bound for the expected contrastive loss to improve the efficiency of our learning algorithm. Thus, graph semantics can be preserved within the augmentations in an intelligent way without arbitrary manual design or prior human knowledge. Experimental results on both graph-level and node-level tasks show that the proposed method achieves state-of-the-art performance compared to other benchmarks, where ablation studies in the end demonstrate the effectiveness of modules in iGCL.