Unlocking the Potential of User Feedback: Leveraging Large Language Model as User Simulator to Enhance Dialogue System

TL;DR

This paper introduces UGRO, a novel method that leverages large language models as user simulators to improve task-oriented dialogue systems, achieving superior results by optimizing smaller models with LLM-generated feedback.

Contribution

The paper proposes UGRO, an innovative approach that combines LLMs as user simulators with smaller TOD models to enhance dialogue response quality.

Findings

UGRO outperforms previous state-of-the-art methods on two TOD benchmarks.

Using LLMs as user simulators effectively guides smaller models to generate higher satisfaction responses.

Empirical results validate the effectiveness of the proposed approach.

Abstract

Dialogue systems and large language models (LLMs) have gained considerable attention. However, the direct utilization of LLMs as task-oriented dialogue (TOD) models has been found to underperform compared to smaller task-specific models. Nonetheless, it is crucial to acknowledge the significant potential of LLMs and explore improved approaches for leveraging their impressive abilities. Motivated by the goal of leveraging LLMs, we propose an alternative approach called User-Guided Response Optimization (UGRO) to combine it with a smaller TOD model. This approach uses LLM as annotation-free user simulator to assess dialogue responses, combining them with smaller fine-tuned end-to-end TOD models. By utilizing the satisfaction feedback generated by LLMs, UGRO further optimizes the supervised fine-tuned TOD model. Specifically, the TOD model takes the dialogue history as input and, with the assistance of the user simulator's feedback, generates high-satisfaction responses that meet the user's requirements. Through empirical experiments on two TOD benchmarks, we validate the effectiveness of our method. The results demonstrate that our approach outperforms previous state-of-the-art (SOTA) results.