Evidential theoretic modeling and analysis of recognition primed cognitive decision processes under incomplete knowledge and information

不完整知识和信息下识别引发的认知决策过程的证据理论建模和分析

• 批准号: 12680366
• 负责人: INAGAKI Toshiyuki
• 金额: $ 2.24万
• 依托单位: University of Tsukuba
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (C)
• 财政年份: 2000
• 资助国家: 日本
• 起止时间: 2000 至 2001
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-12680366/
• 关键词: Theory of evidence; Normative decision-making; Naturalistic decision-making; Risk perception; Reliability and safety; Trust and complacency; Authority; Human-centered automation; 不確実情報; 意思決定; 認知誘導型決定; ヒューマンマシンシステム; 機能配分; 権限委譲; 情報更新; 不完全情報 /

This project has three research goals. The first goal is to develop a theory for adaptive function allocation (or, adaptive automation) between humans and automation with incomplete knowledge and information. Four phases are distinguished for cognitive processes : information acquisition, information analysis, decision-making and action selection, and action implementation. Sharing and trading of control between humans and automation are discussed for each phase.The second research goal is to develop a framework for human decision making under time stress. Some researchers have claimed recently that normative decision models fail to represent expert decisions. This study has proven that the claim may not always be true, and has proposed a continuum of decision problems that ranges from a purely recognition-primed model to a purely normative decision-making for investigation of wide variety of realistic decision problems.The third goal is investigation of belief updating mechanism in human cognitive decision processes. An experimental research vehicle has created and cognitive experiment has been conducted. It has been observed that most subjects updated their belief function values with some mechanism that has a sort of 'inertia' to their previous belief values.

该项目有三个研究目标。第一个目标是开发一种理论，用于在人类和不完全知识和信息的自动化之间进行自适应功能分配（或自适应自动化）。认知过程分为四个阶段：信息获取、信息分析、决策和行动选择、行动实施。第二个研究目标是为人类在时间压力下的决策提供一个框架。一些研究人员最近声称，规范决策模型不能代表专家的决策。本研究证明了这一观点并不总是正确的，并提出了一个从纯认知启动模型到纯规范决策的决策问题连续体，用于研究各种现实决策问题。第三个目标是研究人类认知决策过程中的信念更新机制。建立了一个实验研究车，并进行了认知实验。已经观察到，大多数受试者更新他们的信念函数值与一些机制，有一种“惯性”，他们以前的信念值。

项目成果

T.Inagaki: "Automation and the cost of authority"Int. J. Industrial Ergonomics. (掲載予定).

T. Inagaki：“自动化与权威成本”Int. J. Industrial Ergonomics（即将出版）。

T. Inagaki, R. Parasuraman: "Probabilistic analysis of human interaction with automated alerts"Proc. 5^<th> Conference on Probabilistic Safety Assessment and Management (PSAM). 2405-2410 (2000)

T. Inagaki，R. Parasuraman：“人类与自动警报交互的概率分析”Proc。

T.Inagaki: "Probabilistic analysis of human interaction with automated alerts"Proc.5th Conf.Probabilistic Safety Assessment and Management. 2405-2410 (2000)

T.Inagaki：“人类与自动警报交互的概率分析”Proc.5th Conf.概率安全评估和管理。

T.Inagaki: "Adaptive function allocation: can it contribute to system safety?"Proc. AIR 2002. 164-171 (2002)

T.Inagaki：“自适应功能分配：它有助于系统安全吗？”

T.Inagaki: "Situation-adaptive autonomy : Dynamic trading of authority between human and automation"Proc.IEA2000/HFES2000 Congress. III. 13-16 (2000)

T.Inagaki：“情境自适应自治：人类与自动化之间的动态权力交易”Proc.IEA2000/HFES2000 大会。