SMRT Chatbots: Improving Non-Task-Oriented Dialog with Simulated Multiple Reference Training

TL;DR

This paper introduces SMRT, a training method that uses simulated multiple responses to enhance non-task-oriented chatbots, resulting in improved response quality and diversity without needing additional domain-specific data.

Contribution

The paper presents SMRT, a novel training approach that leverages paraphrasing to simulate multiple responses, significantly improving chatbot response quality and diversity.

Findings

SMRT outperforms a strong Transformer baseline in quality and diversity.

SMRT is comparable to pretraining in human evaluations.

SMRT surpasses pretraining on automatic quality and lexical diversity.

Abstract

Non-task-oriented dialog models suffer from poor quality and non-diverse responses. To overcome limited conversational data, we apply Simulated Multiple Reference Training (SMRT; Khayrallah et al., 2020), and use a paraphraser to simulate multiple responses per training prompt. We find SMRT improves over a strong Transformer baseline as measured by human and automatic quality scores and lexical diversity. We also find SMRT is comparable to pretraining in human evaluation quality, and outperforms pretraining on automatic quality and lexical diversity, without requiring related-domain dialog data.