GraphHAM

This repository contains the source code and dataset for our paper "An Efficient Automatic Meta-Path Selection for Social Event Detection via Hyperbolic Space".

Requirements

Python>=3.7
torch>=1.4.0
scipy>=1.2.1
networkx>=2.4
scikit-learn>=0.20.3

To run GraphHAM

python run.py

Datasets

The datasets used in GraphHAM are Kawarith[1], CrisisLexT6[2], and Twitter2012[3]. To comply with Twitter’s policies, we only uploaded the processed Kawarith dataset for demonstration. You can retrieve the complete tweet object via the dataset's link and a valid Twitter API.

Ablation Studies

All variables can be modified in config.py

Citation

If you find this repository helpful, please consider citing the following paper.

@inproceedings{qiu2024efficient,
  title={An Efficient Automatic Meta-Path Selection for Social Event Detection via Hyperbolic Space},
  author={Qiu, Zitai and Ma, Congbo and Wu, Jia and Yang, Jian},
  booktitle={Proceedings of the ACM on Web Conference 2024},
  pages={2519--2529},
  year={2024}
}

Reference

[1] Alaa Alharbi and Mark Lee. 2021. Kawarith: an Arabic Twitter corpus for crisis events. In Proceedings of the Sixth Arabic Natural Language Processing Workshop. 42–52.

[2] Alexandra Olteanu, Carlos Castillo, Fernando Diaz, and Sarah Vieweg. 2014. Crisislex: A lexicon for collecting and filtering microblogged communications in crises. In Proceedings of the international AAAI conference on web and social media, Vol. 8. 376–385.

[3] Andrew J McMinn, Yashar Moshfeghi, and Joemon M Jose. 2013. Building a large-scale corpus for evaluating event detection on twitter. In Proceedings of the 22nd ACM international conference on Information & Knowledge Management. 409–418.

timothy2327 / GraphHAM