This repository contains the pre-processed data and code for our EMNLP 2018 paper "BanditSum: Extractive Summarization as a Contextual Bandit". Please contact me at [email protected] for any question.

Please cite this paper if you use our code or data.

@inproceedings{dong2018banditsum,
  title={BanditSum: Extractive Summarization as a Contextual Bandit},
  author={Dong, Yue and Shen, Yikang and Crawford, Eric and van Hoof, Herke and Cheung, Jackie Chi Kit},
  booktitle={Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing},
  pages={3739--3748},
  year={2018}
}

Instructions to download our preprocessed CNN/DailyMail Dataset can be found here (coming soon).

Our code requires PyTorch version >= 0.4.0. Please follow the instructions here: https://github.com/pytorch/pytorch#installation.

After PyTorch is installed, you can install run our model through main.py.