Self-supervised Graph-level Representation Learning with Local and Global Structure
It addresses the problem of graph-level representation learning for tasks like molecule properties prediction, offering an incremental improvement by incorporating global structure.
The paper tackles unsupervised whole-graph representation learning by proposing GraphLoG, which preserves local similarities and captures global semantic clusters using hierarchical prototypes, achieving effectiveness in chemical and biological benchmark datasets.
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.