ProGCL: Rethinking Hard Negative Mining in Graph Contrastive Learning

TL;DR

ProGCL introduces a novel negative sampling strategy for graph contrastive learning that estimates the probability of negatives being true negatives, significantly improving performance and achieving state-of-the-art results.

Contribution

The paper proposes ProGCL, a method to better select negatives in graph contrastive learning by estimating their true negative probability, addressing false negatives caused by message passing.

Findings

ProGCL outperforms existing GCL methods on multiple benchmarks.

ProGCL achieves state-of-the-art results, surpassing some supervised methods.

ProGCL can be integrated into various negatives-based GCL techniques.

Abstract

Contrastive Learning (CL) has emerged as a dominant technique for unsupervised representation learning which embeds augmented versions of the anchor close to each other (positive samples) and pushes the embeddings of other samples (negatives) apart. As revealed in recent studies, CL can benefit from hard negatives (negatives that are most similar to the anchor). However, we observe limited benefits when we adopt existing hard negative mining techniques of other domains in Graph Contrastive Learning (GCL). We perform both experimental and theoretical analysis on this phenomenon and find it can be attributed to the message passing of Graph Neural Networks (GNNs). Unlike CL in other domains, most hard negatives are potentially false negatives (negatives that share the same class with the anchor) if they are selected merely according to the similarities between anchor and themselves, which will undesirably push away the samples of the same class. To remedy this deficiency, we propose an effective method, dubbed \textbf{ProGCL}, to estimate the probability of a negative being true one, which constitutes a more suitable measure for negatives' hardness together with similarity. Additionally, we devise two schemes (i.e., \textbf{ProGCL-weight} and \textbf{ProGCL-mix}) to boost the performance of GCL. Extensive experiments demonstrate that ProGCL brings notable and consistent improvements over base GCL methods and yields multiple state-of-the-art results on several unsupervised benchmarks or even exceeds the performance of supervised ones. Also, ProGCL is readily pluggable into various negatives-based GCL methods for performance improvement. We release the code at \textcolor{magenta}{\url{https://github.com/junxia97/ProGCL}}.