libing125 / CTEG

Code for ACL 2019 paper "Enhancing Topic-to-Essay Generation with External Commonsense Knowledge "

CTEG

Commonsense enhanced Topic-to-Essay Generation: code for ACL 2019 paper "Enhancing Topic-to-Essay Generation with External Commonsense Knowledge".

Corresponding codes of MTA baselines can be found at MTA-LSTM.

Model

Requirements

• Python 3
• TensorFlow >= 1.7

Data

Data can be downloaded from Baidu Cloud

Run

1. Unzip the data and place it according to the path in config.py
2. Run python3 cteg.py

Feel free to contact me or raise an issue if you meet any problems.

Citation

If you use this code or the corresponding baseline code, please kindly cite our paper:

@inproceedings{yang2019enhancing,
  title={Enhancing Topic-to-Essay Generation with External Commonsense Knowledge},
  author={Yang, Pengcheng and Li, Lei and Luo, Fuli and Liu, Tianyu and Sun, Xu},
  booktitle={Proceedings of the 57th Conference of the Association for Computational Linguistics},
  pages={2002--2012},
  year={2019}
}