This repository is modified by DIG This is the official implement of Paper On Explainability of Graph Neural Networks via Subgraph Explorations
- clone the repository
- create the env and install the requirements
$ git clone https://github.com/divelab/DIG.git
$ cd DIG/xgraph/SubgraphX
$ source ./install.sh
- Download the required dataset to
DIG/xgraph/dataset
- Download the checkpoints to
DIG/xgraph/SubgraphX/checkpoint
- run the searching scripts with corresponding dataset
$ cd DIG/xgraph/SubgraphX
$ source ./scripts.sh
The hyper-parameters for different models and datasets are shown in this script.
In addition, we also provide the saved searching results.
If you want to reproduce, you can directly download the
results
to DIG/xgraph/SubgraphX/results
Moreover, if you want to train a new model for these datasets, run the training scripts for corresponding dataset.
$ cd DIG/xgraph/SubgraphX
$ source ./models/train_gnns.sh
If you use this code, please cite our papers.
@misc{yuan2021explainability,
title={On Explainability of Graph Neural Networks via Subgraph Explorations},
author={Hao Yuan and Haiyang Yu and Jie Wang and Kang Li and Shuiwang Ji},
year={2021},
eprint={2102.05152},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
@article{yuan2020explainability,
title={Explainability in Graph Neural Networks: A Taxonomic Survey},
author={Yuan, Hao and Yu, Haiyang and Gui, Shurui and Ji, Shuiwang},
journal={arXiv preprint arXiv:2012.15445},
year={2020}
}