ANEMONE

A PyTorch implementation of "ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning", CIKM-21

Dependencies

• python==3.6.1
• dgl==0.4.1
• matplotlib==3.3.4
• networkx==2.5
• numpy==1.19.2
• pyparsing==2.4.7
• scikit-learn==0.24.1
• scipy==1.5.2
• sklearn==0.24.1
• torch==1.8.1

To install all dependencies:

pip install -r requirements.txt

Usage

Anomalies have been injected into three datasets under the ./dataset directory. Please refer to GRAND-Lab/CoLA for graph anomaly injection.

To train and evaluate on Cora:

python run.py --expid 1 --device cuda:0 --runs 1 --alpha 0.8

To train and evaluate on Citeseer:

python run.py --dataset citeseer --expid 2 --device cuda:0 --runs 1 --alpha 0.6

To train and evaluate on Pubmed:

python run.py --dataset pubmed --expid 3 --device cuda:0 --runs 1 --alpha 0.8 --negsamp_ratio_patch 6 --negsamp_ratio_context 1

Citation

If you use our code in your research, please cite the following article:

@inproceedings{jin2021anomaly,
  title={ANEMONE: Graph Anomaly Detection with Multi-Scale Contrastive Learning},
  author={Ming Jin and Yixin Liu and Yu Zheng and Lianhua Chi and Yuan-Fang Li and Shirui Pan},
  booktitle={Proceedings of the 30th ACM International Conference on Information \& Knowledge Management},
  year={2021}
}