First Ask Then Answer: A Framework Design for AI Dialogue Based on Supplementary Questioning with Large Language Models

ACL ARR 2025 May Submission4342 Authors

Abstract: Large Language Models (LLMs) often struggle to deliver accurate and actionable answers when user-provided information is incomplete or ill-specified. We propose a new interaction paradigm, \emph{First Ask Then Answer} (FATA), in which, through prompt words, LLMs are guided to proactively pose multidimensional supplementary questions to users before answering. Then, using the information received by the users and the original questions to jointly construct prompt words for questioning, a high-quality question-and-answer is ultimately achieved.In contrast to existing clarification approaches—such as the CLAM framework oriented to ambiguity and the self-interrogation Self-Ask method—FATA emphasizes completeness (beyond mere disambiguation) and user participation (inviting human input instead of relying solely on model-internal reasoning). It also adopts a single-turn strategy: all clarifying questions are produced at once, thereby reducing dialogue length and improving efficiency. Conceptually, FATA uses the reasoning power of LLMs to scaffold user expression, elevating ordinary users to an expert-level questioning ability. To evaluate FATA, we constructed a multi-domain benchmark and compared it with two controls: a baseline prompt (B-Prompt) and a context-enhanced expert prompt (C-Prompt). Experimental results show that FATA outperforms B-Prompt by 40% in aggregate metrics and exhibits a coefficient of variation 8% lower than C-Prompt, indicating superior stability.

Paper Type: Long

Research Area: Dialogue and Interactive Systems

Research Area Keywords: spoken dialogue systems;human-in-the-loop;knowledge augmented; applications;grounded dialog;

Contribution Types: NLP engineering experiment

Languages Studied: English,Chinese

Submission Number: 4342