发表机构
AI Foundations, Capital One; Emory University; University of Wisconsin–Madison; University of Central Florida（Capital One AI Foundations; 埃默里大学; 威斯康星大学麦迪逊分校; 中佛罗里达大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出交互式环境AREAs-Lab，结合16个公开数据集构建合成基准与AI模拟用户的自动化评估流水线，助力将用户模糊目标转化为AI系统可执行需求。
AI 中文摘要
构建高效的AI系统越来越依赖于编写高质量的任务需求，但用户往往难以清晰表达决定成功的约束条件、偏好以及边缘情况。在AI开发中这一问题尤为突出，因为AI的行为不仅受人类期望的影响，还受数据特征的塑造。我们提出AREAs-Lab，这是一个面向AI系统的AI驱动需求启发式开发的交互式环境。在AREAs-Lab中，助手通过分析底层数据集并提出针对性的澄清问题，迭代完善最初不完整的需求，以挖掘用户的潜在意图。为了系统研究该场景，我们构建了一个基于16个涵盖不同领域和任务类型的公开数据集的合成基准。每个基准实例包含用户画像、完整参考需求以及作为助手起点的刻意欠规格化版本。我们进一步引入了基于AI模拟用户的自动化评估流水线，该流水线仅在被适当提示时才会揭示隐藏信息，实现了对交互式需求启发质量的可扩展且可重复的评估。AREAs-Lab为研究AI助手如何将模糊的用户目标转化为AI系统可执行的需求提供了受控测试平台。
英文摘要
Building effective AI systems increasingly depends on writing high-quality task requirements, yet users often struggle to articulate the constraints, preferences, and edge cases that determine success. This problem is especially acute in AI development, where behavior is shaped not only by human expectations but also by data characteristics. We present AREAs-Lab, an interactive environment for AI-driven Requirement Elicitation for AI systems. In AREAs-Lab, an assistant iteratively refines an initially incomplete requirement by analyzing the underlying dataset and asking targeted clarification questions to uncover the user's latent intent. To study this setting systematically, we construct a synthetic benchmark grounded in 16 public datasets spanning diverse domains and task types. Each benchmark instance includes a user profile, a complete reference requirement, and an intentionally underspecified version that serves as the assistant's starting point. We further introduce an automated evaluation pipeline based on an AI-simulated user that reveals hidden information only when appropriately prompted, enabling scalable and reproducible assessment of interactive elicitation quality. AREAs-Lab provides a controlled testbed for studying how AI assistants can transform vague user goals into actionable requirements for AI systems.
CommentsAccepted to Findings of EMNLP 2026. 45 pages, 13 figures
