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

Minghao Xu, Hang Wang, Bingbing Ni, Hongyu Guo, Jian Tang
Proceedings of the 38th International Conference on Machine Learning, PMLR 139:11548-11558, 2021.

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v139-xu21g, title = {Self-supervised Graph-level Representation Learning with Local and Global Structure}, author = {Xu, Minghao and Wang, Hang and Ni, Bingbing and Guo, Hongyu and Tang, Jian}, booktitle = {Proceedings of the 38th International Conference on Machine Learning}, pages = {11548--11558}, year = {2021}, editor = {Meila, Marina and Zhang, Tong}, volume = {139}, series = {Proceedings of Machine Learning Research}, month = {18--24 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v139/xu21g/xu21g.pdf}, url = {https://proceedings.mlr.press/v139/xu21g.html}, 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.} }
Endnote
%0 Conference Paper %T Self-supervised Graph-level Representation Learning with Local and Global Structure %A Minghao Xu %A Hang Wang %A Bingbing Ni %A Hongyu Guo %A Jian Tang %B Proceedings of the 38th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2021 %E Marina Meila %E Tong Zhang %F pmlr-v139-xu21g %I PMLR %P 11548--11558 %U https://proceedings.mlr.press/v139/xu21g.html %V 139 %X 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.
APA
Xu, M., Wang, H., Ni, B., Guo, H. & Tang, J.. (2021). Self-supervised Graph-level Representation Learning with Local and Global Structure. Proceedings of the 38th International Conference on Machine Learning, in Proceedings of Machine Learning Research 139:11548-11558 Available from https://proceedings.mlr.press/v139/xu21g.html.

Related Material