Improving Response Quality with Backward Reasoning in Open-domain Dialogue Systems

TL;DR

This paper introduces a bidirectional training approach for open-domain dialogue systems that enhances response informativeness and coherence by integrating backward reasoning with forward generation.

Contribution

It proposes a novel bidirectional training method with backward reasoning, improving response quality without relying on external side information.

Findings

Achieves state-of-the-art response quality performance.

Enhances informativeness and coherence of generated responses.

Does not require pre-trained topic models.

Abstract

Being able to generate informative and coherent dialogue responses is crucial when designing human-like open-domain dialogue systems. Encoder-decoder-based dialogue models tend to produce generic and dull responses during the decoding step because the most predictable response is likely to be a non-informative response instead of the most suitable one. To alleviate this problem, we propose to train the generation model in a bidirectional manner by adding a backward reasoning step to the vanilla encoder-decoder training. The proposed backward reasoning step pushes the model to produce more informative and coherent content because the forward generation step's output is used to infer the dialogue context in the backward direction. The advantage of our method is that the forward generation and backward reasoning steps are trained simultaneously through the use of a latent variable to facilitate bidirectional optimization. Our method can improve response quality without introducing side information (e.g., a pre-trained topic model). The proposed bidirectional response generation method achieves state-of-the-art performance for response quality.