STREAM：一个以数据为中心的框架，用于从流媒体中挖掘高价值任务导向对话
STREAM: A Data-Centric Framework for Mining High-Value Task-Oriented Dialogues from Streaming Media
AI总结：
提出STREAM框架，利用流媒体数据合成大规模多领域任务导向对话数据集StreamDial，通过角色构建和对话蓝图结合RAG生成高质量对话，解决数据稀缺问题。
摘要：
垂直领域的大语言模型受到复杂、特定领域任务导向对话稀缺的瓶颈。现有的数据获取管道面临持续的三难困境：专家标注昂贵，真实服务对话受隐私和商业限制，静态语料库很快过时。我们提出Stream，一个以数据为中心的框架，利用公开可用的流媒体（直播和短视频）大规模合成高价值服务对话。Stream从嘈杂的流中挖掘真实的交互信号，并通过将基于角色的个性构建与对话蓝图构建相结合来合成对话；它进一步采用检索增强生成（RAG）来支持知识感知的响应。基于Stream，我们发布了StreamDial，一个覆盖汽车、餐厅和酒店的大规模多领域数据集。StreamDial总共包含87,498个对话会话和1,497,320轮次，平均每个会话17.11轮，各领域规模相当。每个会话被组织为结构化四元组⟨P_u, P_a, B, H⟩，将对话历史与明确的用户/代理角色和对话蓝图配对，捕捉真实服务行为，如需求挖掘、约束冲突、协商和恢复。使用自动评估和下游任务的评估表明，StreamDial在强基线上提高了内在对话质量，使用StreamDial训练的模型在多个骨干网络上改进了对话状态跟踪；我们进一步报告了完整的人工评估集，并在受控训练预算下在Qwen3-8B上实现了令人鼓舞的多语言迁移。数据发布在https://github.com/hitxueliang/DialogDataSetBySTREAM。
英文摘要：
Large language models for vertical domains are bottlenecked by the scarcity of complex, domain-specific task-oriented dialogues. Existing data acquisition pipelines face a persistent trilemma: expert annotation is expensive, real-world service conversations are constrained by privacy and commercial restrictions, and static corpora quickly become temporally stale. We propose Stream, a data-centric framework that leverages publicly available streaming media (live streams and short videos) to synthesize high-value service dialogues at scale. Stream mines authentic interaction signals from noisy streams and synthesizes conversations by integrating role-grounded persona construction with Conversational Blueprint construction; it further adopts retrieval-augmented generation (RAG) to support knowledge-aware responses. Based on Stream, we release StreamDial, a large-scale multi-domain dataset covering Automotive, Restaurant, and Hotel. StreamDial contains 87,498 dialogue sessions and 1,497,320 turns in total, with an average of 17.11 turns per session and a comparable scale across domains. Each session is organized as a structured quadruplet $\langle P_u, P_a, B, H \rangle$ that pairs dialogue history with explicit user/agent personas and a Conversational Blueprint, capturing realistic service behaviors such as requirement mining, constraint conflicts, negotiation, and recovery. Evaluations with automatic judges and downstream tasks show that StreamDial improves intrinsic dialogue quality over strong baselines, and models trained with StreamDial improve Dialogue State Tracking across backbones; we further report a completed human-evaluation set and encouraging multilingual transfer on Qwen3-8B under a controlled training budget. The data is released in https://github.com/hitxueliang/DialogDataSetBySTREAM.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
