/TKBC

An Interpretable Multi-hop Reasoning (IMR) model for temporal KG forecasting.

Primary LanguagePythonApache License 2.0Apache-2.0

TKBC-IMR

An Interpretable Multi-hop Reasoning (IMR) model for temporal KG forecasting. image

This is the code of paper Interpretable Multi-hop Reasoning for Forecasting Future Links on Temporal Knowledge Graphs.

Dependencies

    pip install -r requirements.txt

Results

  • Dataset in TITer image

  • The result image

Reproduce the Results

    bash ./run_ICEW18.sh
    bash ./run_ICEWS14.sh
    bash ./run_WIKI.sh
    bash ./run_YAGO.sh

Citation

If you find this code useful, please consider citing the following paper.

@inproceedings{
anonymous2022interpretable,
title={Interpretable Multi-hop Reasoning for Forecasting Future Links on Temporal Knowledge Graphs},
author={Anonymous},
booktitle={Submitted to The Thirty-ninth International Conference on Machine Learning},
year={2022},
note={under review}
}

If you have any questions, please email me.

Acknowledgement

We refer to the code of xERTE. Thanks for their contributions.