zeroxleo / HyperGT

The implementation of ICASSP 2024 lecture presentation paper "Hypergraph Transformer for Semi-Supervised Classification"

Hypergraph Transformer for Semi-Supervised Classification

The official implementation for "Hypergraph Transformer for Semi-Supervised Classification" which is accepted to ICASSP 2024 as a lecture presentation. [paper] [slides]

Recommend Environment:

conda create -n "hyperht" python=3.9
conda activate hypergt
bash install.sh

Data Preparation:

Create a folder ../data/raw_data and download the raw datasets from here.

The directory structure should look like:

HyperGT/
  <source code files>
  ...
  data
    raw_data
      congress-bills
      senate-committees
      walmart-trips
      house-committees

Acknowledgement

The pipeline for training is developed on basis of the Nodeformer work. The pipeline for data preprocessing is based on the AllSet work. Sincere appreciation is extended for their valuable contributions.

Citation

If you use this code, please cite our paper:

@inproceedings{liu2023hypergraph,
  title={Hypergraph Transformer for Semi-Supervised Classification},
  author={Zexi Liu and Bohan Tang and Ziyuan Ye and Xiaowen Dong and Siheng Chen and Yanfeng Wang},
  booktitle={ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
  year={2024},
  organization={IEEE}
}