场景知识驱动的即时辅助流水线
A Scenario-Knowledge-Driven Pipeline for Just-in-Time Assistance
- Graduate School of Engineering, The University of Tokyo（东京大学工学部）
- Research into Artifacts, Center for Engineering (RACE)（工程造物研究中心）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
提出一种场景知识驱动的流水线，通过人工编写的场景知识文档配置感知、约束LLM推理并生成分级干预建议，经离线回放验证，实现了可核查的报告和可衡量的升级。
AI中文摘要：
检测到默默挣扎的自助服务终端用户只是第一步；决定是否、何时以及如何提供帮助，依赖于通常深埋于模型权重和阈值中的场景知识。我们提出了一种场景知识驱动的流水线：一份由人工编写并受版本控制的单一场景知识文档，用于配置感知、约束大语言模型推理，并形成分级干预建议。叙述、辅助需求评估和建议相互分离，以便独立审计。作为概念验证，我们离线回放了两段录制的自助服务终端会话，这两段会话是在运行前根据其挣扎证据和回顾性细节选定的。两个案例都支持该设计所承诺的内容：可核查的报告和可衡量的升级。在95次更新中，仅追加叙述的每一句话都引用了其背后的原始事件，规则层在严格标准下检测到13个中的12个和7个中的7个标注的挣扎片段。评估器在恢复时降级，并在证据最大程度收敛时恰好达到最高层级一次。在决定性的求助转折点，叙述、评估和参与者的回顾性描述相互吻合。这些干预措施的适当性、流水线对无挣扎会话的克制，以及文档向新场景框架的迁移，构成了未来的议程。
英文摘要：
Detecting a silently struggling kiosk user is only the first step; deciding whether, when, and how to help depends on scenario knowledge usually buried in model weights and thresholds. We propose a scenario-knowledge-driven pipeline: a single scenario knowledge document, human-authored and version-controlled, configures sensing, constrains LLM reasoning, and shapes a graded intervention proposal. Narration, assistance-need assessment, and proposal are kept separate for independent audit. As proof of concept, we replay two recorded kiosk sessions offline, chosen before the runs for their struggle evidence and retrospective detail. Both cases support what the design promises: checkable reporting and measured escalation. Across 95 updates, every sentence of the append-only narration cites the primitive events underlying it, and the rule layer detects 12 of 13 and 7 of 7 annotated struggle episodes under a strict criterion. The assessor de-escalates on recovery and reaches the top rung exactly once, under maximally converging evidence. At the decisive help-seeking turn, narration, assessment, and the participants' retrospective accounts converge. The appropriateness of these interventions, the pipeline's restraint on sessions without struggle, and the document's transfer to a new scenario frame the agenda.
