Self-supervised Graph-level Representation Learning with Local and Global Structure

TL;DR

This paper introduces GraphLoG, a self-supervised graph representation learning framework that captures both local similarities and global semantic structures using hierarchical prototypes and an EM algorithm, improving performance on chemical and biological tasks.

Contribution

The paper proposes a novel unified framework, GraphLoG, that incorporates hierarchical prototypes and an EM algorithm for global and local structure preservation in graph learning.

Findings

Effective on chemical and biological datasets

Outperforms existing local-only methods

Captures global semantic clusters successfully

Abstract

This paper studies unsupervised/self-supervised whole-graph representation learning, which is critical in many tasks such as molecule properties prediction in drug and material discovery. Existing methods mainly focus on preserving the local similarity structure between different graph instances but fail to discover the global semantic structure of the entire data set. In this paper, we propose a unified framework called Local-instance and Global-semantic Learning (GraphLoG) for self-supervised whole-graph representation learning. Specifically, besides preserving the local similarities, GraphLoG introduces the hierarchical prototypes to capture the global semantic clusters. An efficient online expectation-maximization (EM) algorithm is further developed for learning the model. We evaluate GraphLoG by pre-training it on massive unlabeled graphs followed by fine-tuning on downstream tasks. Extensive experiments on both chemical and biological benchmark data sets demonstrate the effectiveness of the proposed approach.