This is the source code of NeurIPS'23 paper "Towards Self-Interpretable Graph-Level Anomaly Detection" (SIGNET).

Usage

Step 1: prepare datasets

• Mutag:

1. Raw data files need to be downloaded at: https://github.com/flyingdoog/PGExplainer/tree/master/dataset
2. Unzip Mutagenicity.zip and Mutagenicity.pkl.zip
3. Put the raw data files in ./data/mutag/raw

• MNIST:

1. Raw data files need to be generated following the instructions at: https://github.com/bknyaz/graph_attention_pool/blob/master/scripts/mnist_75sp.sh
2. Put the generated files in ./data/mnist/raw

• Others: Download and process automatically

Step 2: run script line in scripts.sh

For example:

python main.py --dataset AIDS --epoch 1000 --lr 0.0001 --hidden_dim 16

Cite

If you compare with, build on, or use aspects of SIGNET, please cite the following:

@inproceedings{liu2023towards,
  title={Towards self-interpretable graph-level anomaly detection},
  author={Liu, Yixin and Ding, Kaize and Lu, Qinghua and Li, Fuyi and Zhang, Leo Yu and Pan, Shirui},
  booktitle={Advances in Neural Information Processing Systems},
  volume={36},
  year={2023}
}