/pos_adv

[NAACL 2018] Robust Sequence Labeling with Adversarial Training

Primary LanguagePythonApache License 2.0Apache-2.0

BiLSTM-CRF Sequence Labeling with Adversarial Training

Implementation of BiLSTM-CRF model with adversarial training.
Paper: Robust Multilingual Part-of-Speech Tagging via Adversarial Training (NAACL 2018).

Requirements

  • Python 2.7
  • Theano 1.0
  • Lasagne

Data

Run

Configure and run multi_lingual_run_blstm-blstm-crf_pos.sh.

Notes

If you use this tool for your work, please consider citing:

@InProceedings{Yasunaga&al.18.naacl,
  author =  {Michihiro Yasunaga and Jungo Kasai and Dragomir R. Radev},
  title =   {Robust Multilingual Part-of-Speech Tagging via Adversarial Training},
  year =    {2018},  
  booktitle =   {Proceedings of NAACL},  
  publisher =   {Association for Computational Linguistics},
}

Acknowledgements

This tool uses the following open source component (big thank you to the developers). You can find its source code and license information below.