Code for "Zero-shot Generalist Graph Anomaly Detection with Unified Neighborhood Prompts"(UMPrompt)

Get Started

To run the code, the following packages are required to be installed:

-python==3.8.19

-torch==1.13.1

-dgl==1.0.1+cu117

Train

To get the results with Facebook as the training graph, just run the following command:

 sh run.sh

Citation

Please acknowledge our work via the following bibliography if you find our work/code useful:

@article{niu2024zero,
  title={Zero-shot Generalist Graph Anomaly Detection with Unified Neighborhood Prompts},
  author={Niu, Chaoxi and Qiao, Hezhe and Chen, Changlu and Chen, Ling and Pang, Guansong},
  journal={arXiv preprint arXiv:2410.14886},
  year={2024}
}