Retrieval-Enhanced Adversarial Training for Neural Response Generation
This work addresses the challenge of improving dialogue systems by combining generation and retrieval approaches, though it appears incremental in nature.
The paper tackles the problem of neural response generation in dialogue systems by proposing a Retrieval-Enhanced Adversarial Training (REAT) method, which significantly outperforms vanilla Seq2Seq and conventional adversarial training on a large-scale benchmark dataset.
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.