Graph Rationalization with Environment-based Augmentations
Authors:
Open Access
- 14 August 2022
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1069-1078
Abstract
Rationale is defined as a subset of input features that best explains or supports the prediction by machine learning models. Rationale identification has improved the generalizability and interpretability of neural networks on vision and language data. In graph applications such as molecule and polymer property prediction, identifying representative subgraph structures named as graph rationales plays an essential role in the performance of graph neural networks. Existing graph pooling and/or distribution intervention methods suffer from the lack of examples to learn to identify optimal graph rationales. In this work, we introduce a new augmentation operation called environment replacement that automatically creates virtual data examples to improve rationale identification. We propose an efficient framework that performs rationale-environment separation and representation learning on the real and augmented examples in latent spaces to avoid the high complexity of explicit graph decoding and encoding. Comparing against recent techniques, experiments on seven molecular and four polymer datasets demonstrate the effectiveness and efficiency of the proposed augmentation-based graph rationalization framework. Data and the implementation of the proposed framework are publicly available https://github.com/liugangcode/GREA.Keywords
All Related Versions
Funding Information
- NSF (National Science Foundation) (1849816, 2142827, 2146761, 2102592)
This publication has 18 references indexed in Scilit:
- Polymer informatics: Current status and critical next stepsMaterials Science and Engineering: R: Reports, 2021
- Data Augmentation for Graph Classification2020
- PI1M: A Benchmark Database for Polymer InformaticsJournal of Chemical Information and Modeling, 2020
- Calendar Graph Neural Networks for Modeling Time Structures in Spatiotemporal User Behaviors2020
- NodeAug: Semi-Supervised Node Classification with Data Augmentation2020
- Measuring and Relieving the Over-Smoothing Problem for Graph Neural Networks from the Topological ViewProceedings of the AAAI Conference on Artificial Intelligence, 2020
- Evaluating Polymer Representations via Quantifying Structure–Property RelationshipsJournal of Chemical Information and Modeling, 2019
- Polymer Genome: A Data-Powered Polymer Informatics Platform for Property PredictionsThe Journal of Physical Chemistry C, 2018
- Energy-efficient polymeric gas separation membranes for a sustainable future: A reviewPolymer, 2013
- PoLyInfo: Polymer Database for Polymeric Materials Design2011