This repository contains the source code and datasets for the IR Journal paper ReBoost: A Retrieval-Boosted Sequence-to-Sequence Model for Neural Response Generation by Zhu et al.
Python 3.5
Tensorflow 1.1.2
Your can download the processed datasets used in our paper here and unzip it to the folder of data.
Weibo (STC)
OpenSubtitle
python3 runModel.py
If you use the code and datasets, please cite the following paper:
"ReBoost: A Retrieval-Boosted Sequence-to-Sequence Model for Neural Response Generation"
Yutao Zhu, Zhicheng Dou, Jian-Yun Nie and Ji-Rong Wen. IR Journal (2019)
@article{DBLP:journals/ir/ZhuDNW20,
author = {Yutao Zhu and
Zhicheng Dou and
Jian{-}Yun Nie and
Ji{-}Rong Wen},
title = {ReBoost: a retrieval-boosted sequence-to-sequence model for neural
response generation},
journal = {Inf. Retr. Journal},
volume = {23},
number = {1},
pages = {27--48},
year = {2020},
url = {https://doi.org/10.1007/s10791-019-09364-x},
doi = {10.1007/s10791-019-09364-x}
}