Dialogue systems are usually built on either generation-based or retrieval-based approaches, yet they do not benefit from the advantages of different models. In this paper, we propose a Retrieval-Enhanced Adversarial Training (REAT) method for neural response generation. Distinct from existing approaches, the REAT method leverages an encoder-decoder framework in terms of an adversarial training paradigm, while taking advantage of N-best response candidates from a retrieval-based system to construct the discriminator. An empirical study on a large scale public available benchmark dataset shows that the REAT method significantly outperforms the vanilla Seq2Seq model as well as the conventional adversarial training approach.
@article{arxiv.1809.04276,
title = {Retrieval-Enhanced Adversarial Training for Neural Response Generation},
author = {Qingfu Zhu and Lei Cui and Weinan Zhang and Furu Wei and Ting Liu},
journal= {arXiv preprint arXiv:1809.04276},
year = {2019}
}