Cue-CoT: Chain-of-thought Prompting for Responding to In-depth Dialogue Questions with LLMs
This addresses the problem of generating personalized and engaging responses in dialogue systems for users, but it is incremental as it builds on existing chain-of-thought methods by focusing on specific linguistic cues.
The paper tackles the challenge of LLMs responding to in-depth dialogue questions by overlooking linguistic cues like personality, emotion, and psychology, proposing Cue-CoT, a chain-of-thought prompting method that enhances inference with an intermediate reasoning step, resulting in improved helpfulness and acceptability across 6 datasets in Chinese and English.
Large Language Models (LLMs), such as \texttt{ChatGPT}, greatly empower dialogue systems with strong language understanding and generation capabilities. However, most of the previous works prompt the LLMs to directly generate a response based on the dialogue context, overlooking the underlying linguistic cues about the user status exhibited in the context. Such in-depth dialogue scenarios are challenging for existing LLMs to figure out the user's hidden needs and respond satisfactorily through a single-step inference. To this end, we propose a novel linguistic cue-based chain-of-thoughts (\textit{Cue}-CoT), which enhances the LLMs inference with an intermediate reasoning step to find cues exhibited in the dialogue, aiming to provide a more personalized and engaging response. To evaluate the approach, we build a benchmark with in-depth dialogue questions, consisting of 6 datasets in both Chinese and English, targeting 3 major linguistic cues during the conversation: \textit{personality}, \textit{emotion}, and \textit{psychology}. We conduct extensive experiments on the proposed benchmark with 5 LLMs under both zero-shot and one-shot settings. Empirical results demonstrate our proposed \textit{Cue}-CoT method outperforms standard prompting methods in terms of both \textit{helpfulness} and \textit{acceptability} on all datasets.