saltwater-fish / dcrs

DCRS: Encoding Node Diffusion Competence and Role Significance for Network Dismantling

Encoding Node Diffusion Competence and Role Significance for Network Dismantling

The official PyTorch implementation of DCRS in the following paper:

Jiazheng Zhang, Bang Wang. 2023. Encoding Node Diffusion Competence and Role Significance for Network Dismantling.
In WWW'23, April 30 – May 4, 2023, Austin, Texas, USA, 12 pages.

Dependencies

• torch 1.7.1
• igraph 0.9.9
• networkx 2.6.3
• scipy 1.5.4
• sklearn 0.24.2
• numpy 1.19.5
• dgl 0.7.0

Install all dependencies using

pip install -r requirement.txt

Usage

1. Unzip the data.zip and put them in the folder .\data .
2. Train and evaluate the model on real-world networks :

python train_realworld.py

3. Train and evaluate the model on synthetic networks :

python train_synthetic.py --Type BA --N 1000 --M 4

--Type argument should be one of [ BA, WS, PLC, ER], --N set the size of synthetic networks, --M set the synthetic generation parameter.

Citation

Please cite our work if you find our code/paper is helpful to your work.

@inproceedings{zhang2023DCRS,
  title={Encoding Node Diffusion Competence and Role Significance for Network Dismantling},
  author={Zhang, Jiazheng and Wang, Bang},
  booktitle={Proceedings of the 2023 ACM Web Conference},
  series={WWW'23},
  year={2023},
  location={Austin, Texas, USA},
  numpages={12}
}