yangzhou12 / NCRL

Implementation of None Class Ranking Loss for Document-Level Relation Extraction

NCRL

This repository is the implementation of None Class Ranking Loss for Document-Level Relation Extraction (IJCAI 2022).

We make the NCRL loss function public. The experimental results on DocRED can be reproduced by replacing the loss function of ATLOP with NCRL.

Citation

Please cite the following paper if you find our code is useful. Thanks!

@inproceedings{zhou2022none,
  title={None Class Ranking Loss for Document-Level Relation Extraction},
  author={Yang Zhou, Wee Sun Lee},
  booktitle={Proceedings of IJCAI},
  year={2022}
}