Leveraging Adversarial Training in Self-Learning for Cross-Lingual Text Classification

This is Pytorch implemenation of our papers: A Robust Self-Learning Framework for Cross-Lingual Text Classification, Leveraging Adversarial Training in Self-Learning for Cross-Lingual Text Classification, introduced by Xin Dong, Gerard de Melo.

Requirements

python 3.7
pytorch 1.0

Datasets

This sample code uses MLDoc dataset which is not public because of the privacy. You can request it on your own. Besides, you can create new data_processor for you own dataset in run_ld.py.

Training

Please run

./run_ld.sh

To run adversarial training, Plese add

--adv_training 

in run_ld.sh

Citation

If you found this code/work to be useful in your own research, please considering citing the following:

@inproceedings{dong2019robust,
  title={A robust self-learning framework for cross-lingual text classification},
  author={Dong, Xin and de Melo, Gerard},
  booktitle={Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)},
  pages={6307--6311},
  year={2019}
}

@inproceedings{dong2020leveraging,
  title={Leveraging Adversarial Training in Self-Learning for Cross-Lingual Text Classification},
  author={Dong, Xin and Zhu, Yaxin and Zhang, Yupeng and Fu, Zuohui and Xu, Dongkuan and Yang, Sen and de Melo, Gerard},
  booktitle={Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval},
  pages={1541--1544},
  year={2020}
}