A Preliminary Evaluation of ChatGPT for Zero-shot Dialogue Understanding

TL;DR

This paper evaluates ChatGPT's zero-shot dialogue understanding capabilities, demonstrating its potential and limitations across various benchmarks and tasks, and providing insights for future LLM-based dialogue systems.

Contribution

It is the first comprehensive assessment of ChatGPT's zero-shot performance on dialogue understanding, highlighting its strengths and weaknesses in different tasks.

Findings

ChatGPT shows strong potential in zero-shot dialogue understanding.

Multi-turn prompts improve dialogue state tracking performance.

ChatGPT struggles with slot filling in spoken language understanding.

Abstract

Zero-shot dialogue understanding aims to enable dialogue to track the user's needs without any training data, which has gained increasing attention. In this work, we investigate the understanding ability of ChatGPT for zero-shot dialogue understanding tasks including spoken language understanding (SLU) and dialogue state tracking (DST). Experimental results on four popular benchmarks reveal the great potential of ChatGPT for zero-shot dialogue understanding. In addition, extensive analysis shows that ChatGPT benefits from the multi-turn interactive prompt in the DST task but struggles to perform slot filling for SLU. Finally, we summarize several unexpected behaviors of ChatGPT in dialogue understanding tasks, hoping to provide some insights for future research on building zero-shot dialogue understanding systems with Large Language Models (LLMs).