Skip to content
/ DocuNet Public

[IJCAI 2021] Document-level Relation Extraction as Semantic Segmentation

License

Notifications You must be signed in to change notification settings

zjunlp/DocuNet

Folders and files

NameName
Last commit message
Last commit date

Latest commit

aa946e1 · Dec 6, 2022

History

64 Commits
Aug 31, 2021
Aug 31, 2021
Aug 31, 2021
Aug 31, 2021
May 1, 2022
Aug 31, 2021
Apr 20, 2022
May 7, 2021
Dec 6, 2022
Aug 31, 2021
Aug 31, 2021
Sep 29, 2021
Aug 31, 2021
Aug 31, 2021
Oct 20, 2022
Aug 31, 2021
Aug 31, 2021
Sep 29, 2021
Aug 31, 2021
Aug 31, 2021
Oct 31, 2021
Dec 28, 2021
Sep 13, 2021

Repository files navigation

DocuNet

This repository is the official implementation of DocuNet, which is model proposed in a paper: Document-level Relation Extraction as Semantic Segmentation, accepted by IJCAI2021 main conference.

  • ❗NOTE: Docunet is integrated in the knowledge extraction toolkit DeepKE.

Brief Introduction

This paper innovatively proposes the DocuNet model, which first regards the document-level relation extraction as the semantic segmentation task in computer vision.

Requirements

To install requirements:

pip install -r requirements.txt

Training

To train the DocuNet model in the paper on the dataset DocRED, run this command:

>> bash scripts/run_docred.sh # use BERT/RoBERTa by setting --transformer-type

To train the DocuNet model in the paper on the dataset CDR and GDA, run this command:

>> bash scripts/run_cdr.sh  # for CDR
>> bash scripts/run_gda.sh  # for GDA

Evaluation

To evaluate the trained model in the paper, you setting the --load_path argument in training scripts. The program will log the result of evaluation automatically. And for DocRED it will generate a test file result.json in the official evaluation format. You can compress and submit it to Colab for the official test score.

Results

Our model achieves the following performance on :

Document-level Relation Extraction on DocRED

Model Ign F1 on Dev F1 on Dev Ign F1 on Test F1 on Test
DocuNet-BERT (base) 59.86±0.13 61.83±0.19 59.93 61.86
DocuNet-RoBERTa (large) 62.23±0.12 64.12±0.14 62.39 64.55

Document-level Relation Extraction on CDR and GDA

Model CDR GDA
DocuNet-SciBERT (base) 76.3±0.40 85.3±0.50

Acknowledgement

Part of our code is borrowed from https://github.com/wzhouad/ATLOP, many thanks. You can refer to https://github.com/fenchri/edge-oriented-graph for the detailed preprocessing process of GDA and CDR datasets (acquire the file of train_filter.data, dev_filter.data and test_filter.data).

Papers for the Project & How to Cite

If you use or extend our work, please cite the paper as follows:

@inproceedings{ijcai2021-551,
  title     = {Document-level Relation Extraction as Semantic Segmentation},
  author    = {Zhang, Ningyu and Chen, Xiang and Xie, Xin and Deng, Shumin and Tan, Chuanqi and Chen, Mosha and Huang, Fei and Si, Luo and Chen, Huajun},
  booktitle = {Proceedings of the Thirtieth International Joint Conference on
               Artificial Intelligence, {IJCAI-21}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {Zhi-Hua Zhou},
  pages     = {3999--4006},
  year      = {2021},
  month     = {8},
  note      = {Main Track}
  doi       = {10.24963/ijcai.2021/551},
  url       = {https://doi.org/10.24963/ijcai.2021/551},
}