Leveraging LLMs for Dialogue Quality Measurement
This work addresses the challenge of poor correlation with human judgments in task-oriented conversational AI evaluation, offering an incremental improvement for researchers and developers in NLP.
The paper tackled the problem of automated dialogue quality evaluation by exploring the use of large language models (LLMs) with various configurations, finding that fine-tuned LLMs with chain-of-thought reasoning and algorithmic example selection improve performance.
In task-oriented conversational AI evaluation, unsupervised methods poorly correlate with human judgments, and supervised approaches lack generalization. Recent advances in large language models (LLMs) show robust zeroshot and few-shot capabilities across NLP tasks. This paper explores using LLMs for automated dialogue quality evaluation, experimenting with various configurations on public and proprietary datasets. Manipulating factors such as model size, in-context examples, and selection techniques, we examine "chain-of-thought" (CoT) reasoning and label extraction procedures. Our results show that (1) larger models yield more accurate dialogue labels; (2) algorithmic selection of in-context examples outperforms random selection; (3) CoT reasoning where an LLM is asked to provide justifications before outputting final labels improves performance; and (4) fine-tuned LLMs outperform out-of-the-box ones. Our results indicate that LLMs that are suitably fine-tuned and have sufficient reasoning capabilities can be leveraged for automated dialogue evaluation.