News (June 27, 2025):

I am so glad to share that the advanced version of this model (with PyTorch) has been implemented and now integrated into the well-known hypergraph library: EasyGraph

You can directly access the source code of HWNN in the repository Here HWNN

or you can use it in the EasyGraph framework via its sub-area tool: EasyHypergraph.

easygraph.model.hypergraphs.hwnn

In the near future, I will try to finish the code tutorial, usage example, etc to make this work impact more widely. (If you could help me finish this work, it would be greatly appreciated!)

Please cite our paper if possible:

@inproceedings{sun2020hwnn,
  title={Heterogeneous Hypergraph Embedding for Graph Classification},
  author={Sun, Xiangguo and 
  Yin, Hongzhi and 
  Liu, Bo and 
  Chen, Hongxu and 
  Shao, Yingxia and 
  Viet Hung, Nguyen Quoc},
  booktitle={14th ACM International Conference on Web Search and Data Mining (WSDM2021)},
  year={2021}
}

History Message:

This is the source code (beta version) of our paper: Xiangguo Sun et al. Heterogeneous Hypergraph Embedding for Graph Classification, WSDM2021

A more advanced version will be released in the near future (around January, 2021).

Datasets: Pumbed, Cora, DBLP are included in the folder 'data'

For the Spammer dataset, please contact the authors of the following paper to obtain the permission: Bo Liu et al. Co-Detection of Crowdturfing Microblogs and Spammers in Online Social Networks. World Wide Web Journal (WWWJ). 2020, 23, 573–607