Awesome

DynamicGCN

This is the source code for paper Learning Dynamic Context Graphs for Predicting Social Events appeared in KDD2019 (research track)

Songgaojun Deng, Huzefa Rangwala, Yue Ning

Data

• ICEWS event data is available online.
• ICEWS news data has not been released publicly.（If you want to access the original news text information of the event, I suggest GDELT data.）

Libraries

• PyTorch >= 1.0
• sklearn
• pytorch_sparse Refer to the official website to install.

Sample dataset

• THAD6h (Thailand dynamic (temporal) dataset, around 600 nodes per graph) Google Drive
• INDD6h Google Drive
• EGYD6h Google Drive
• RUSD6h Google Drive
  • *.idx / *.tidx Word index file for training/testing
  • *.x / *.tx Temporal graph input file for training/testing
  • *.y / *.ty Ground truth for training/testing

Cite

Please cite our paper if you find this code useful for your research:

@inproceedings{deng2019learning,
  title={Learning Dynamic Context Graphs for Predicting Social Events},
  author={Deng, Songgaojun and Rangwala, Huzefa and Ning, Yue},
  booktitle={Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining},
  pages={1007--1016},
  year={2019}
}