GraphCL: Contrastive Self-Supervised Learning of Graph Representations
This work addresses the challenge of unsupervised representation learning for graph-structured data, which is crucial for tasks like node classification in domains such as social networks or bioinformatics, and it presents a novel method that advances the field beyond existing approaches.
The paper tackles the problem of learning node representations in graphs without supervision by proposing GraphCL, a contrastive self-supervised framework that maximizes similarity between perturbed versions of node subgraphs, achieving significant improvements over state-of-the-art methods on node classification benchmarks.
We propose Graph Contrastive Learning (GraphCL), a general framework for learning node representations in a self supervised manner. GraphCL learns node embeddings by maximizing the similarity between the representations of two randomly perturbed versions of the intrinsic features and link structure of the same node's local subgraph. We use graph neural networks to produce two representations of the same node and leverage a contrastive learning loss to maximize agreement between them. In both transductive and inductive learning setups, we demonstrate that our approach significantly outperforms the state-of-the-art in unsupervised learning on a number of node classification benchmarks.