透過您的圖書館登入
IP:216.73.216.60
  • 學位論文

人工智能應用機會的系統化辨識手法

Systematic Identification of Artificial Intelligence Application Opportunities

指導教授 : 許棟樑 葉維彰

摘要


目前對人工智慧(AI)應用機會的辨識,普遍是由專家的腦力激盪產生,高度倚賴運氣和靈感。本研究發展了一套人工智能應用機會的系統化辨識手法,該流程利用情境分析和IDEF建模方法來識別和表示現有流程場景,並通稱化/價值化基本需求要素,簡化/合理化、自動化、智能化現有情境,最後,找出流程中有哪些智能化機會,根據需求點對流程進行智能化創新,設計根據人工智能功能要素的啟發示問句辨識情境中是否有人工智能應用機會,並使用修改後的IDEF模型呈現新的創新機會與條列出功能設備需求規格,令後續專家與學界能運用此需求表,提供企業和研究支持與服務。 本研究貢獻含:(1) 使用通稱化/價值化後的詞語來跳出思維窠臼,以識別情境中物件的潛在替代元件/創新機會; (2)發展一套IDEF裁剪方法運用於流程簡化/合理化,使流程在自動化以前能避免不必要的浪費與錯誤; (3)提出智能化功能要素啟發示問句,能辨識現有情境中的潛在人工智能應用機會; (4)整理出完整流程、表單、說明,並引入案例貫穿整體流程以利實務上之應用。

並列摘要


So far, the identification of artificial intelligence (AI) application opportunities is generally produced by brainstorming of experts and is highly relying on luck and inspiration. This research developed a set of systematic processes to facilitate systematic identification of AI application opportunities. The process utilizes scenario analysis and IDEF modeling to identify and represent existing process scenarios. Generalization of objects and terminology, rationalization of the processes, automation/digitalization, and matching of AI capability elements with scenario details are used to facilitate identification of AI application opportunities. Thought-provoking questions are designed to match AI capability elements and process details so as to identify potential usage of AI capabilities. The identified application scenarios are represented by revised IDEF model to illustrtate elements of the new scenario which helps the users to identify high level requirements of the AI applications. The developers of the AI applications can use the high level requirements to develop the ensuing applications. The contributions of this research include: (1) using object and terminology generalization to jump out of thinking box to identify potential alternatives for current objects in the scenario; (2) developing methods of process trimming to rationalize processes and remove unnecessary processes before digitalization and usage of AI elements; (3) Developing throught-provoking questions to identify potentially applicable AI capability element for scenario usage; (4) Providing forms, instructions, and cases to facilitate practical usage of the method.

參考文獻


中文參考文獻
劉大綺(2019),軟體及流程方法群組專利迴避手法,國立清華大學工業工程研究所碩士論文,新竹。
羅允良(2018),整合性創新產品機會分析與定義流程,國立清華大學工業工程研究所碩士論文,新竹。
陳則翰(2018),問題重框架以解題:系統化的系統轉移及因果分析,國立清華大學工業工程研究所碩士論文,新竹。
林奕宏(2018),專利強化手法以辨識新產品及概念設計,國立清華大學工業工程研究所碩士論文,新竹。

延伸閱讀