DiscProReco

This repository contains the source code of our paper, "A Joint Model for Dropped Pronoun Recovery and Conversational Discourse Parsing in Chinese Conversational Speech", which is accepted for publication at ACL 2021.

Details

1. train_v4.py: the training process of our model
2. model_v4.py: the framework of our DiscProReco
3. results of SPDPR.txt: the results of DiscProReco on SPDPR dataset.

Usage

python train_v4.py 

Datasets

Citation

If this work is useful in your research, please kindly cite our paper.