StevenyzZhang/Guided-Adversarial-Augmentation

Guided-Adversarial-Augmentation

Code and Dataset for the paper:

Aaron Reich, Jiaao Chen, Aastha Agrawal, Yanzhe Zhang, Diyi Yang: Leveraging Expert Guided Adversarial Augmentation to Improve Generalization in Named Entity Recognition, ACL 2022 (Findings).

If you would like to refer to it, please cite the paper mentioned above (Arxiv).

@misc{reich2022leveraging,
    title={Leveraging Expert Guided Adversarial Augmentation For Improving Generalization in Named Entity Recognition},
    author={Aaron Reich and Jiaao Chen and Aastha Agrawal and Yanzhe Zhang and Diyi Yang},
    year={2022},
    eprint={2203.10693},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

Requirements

python >= 3.6

pytorch >= 1.4.0

transformers

numpy

Data

Please unzip the data.zip file. It contains within it the Challenge Set. The Challenge Set’s examples are annotated with a 1 for high quality and a 0 or 2 for low quality. The code only reads in examples annotated with a 1 from the file.

Code

Please run the commands contained in the "README Commands" file for data processing and reproduction of experiments.