PersuasiveDialogue

Code of LREC-COLING 2024 long paper "Would You Like to Make a Donation? A Dialogue System to Persuade You to Donate"

Data

The PersuasionForGood dataset is from Wang et al., 2019. Persuasion for Good: Towards a Personalized Persuasive Dialogue System for Social Good.. You can refer to it for more details about the dataset.

Run

1. Run preprocess.py to preprocess the dataset
2. Run main_bert.py / multilabel.py to train and test a Persuasion Strategy Selection Module.
3. Run train_gpt.py to train a Natural Language Generation Module
4. Run persuasiveness.py to train and test a Persuasiveness Prediction Model
5. Run eval.py to evaluate the Natural Language Generation Module
6. Run interact.py to combine the Persuasion Strategy Selection Module and the Natural Language Generation Module into a persuasive dialogue system, and then you can talk to it!

Citation

If you use any source codes included in this repository in your work, please cite our paper:

@inproceedings{song2024would,
    title={Would You Like to Make a Donation? A Dialogue System to Persuade You to Donate},
    author={Song, Yuhan and Wang, Houfeng},
    booktitle={Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)},
    pages={17707--17717},
    year={2024}
}