許永真臺灣大學：資訊工程學研究所鄭紹佑Cheng, Shao-youShao-youCheng2007-11-262018-07-052007-11-262018-07-052007http://ntur.lib.ntu.edu.tw//handle/246246/53975在普及運算（Pervasive Computing）科技到來的時代，智慧型系統的 設計往往取決於它是否能夠感知周圍環境的變化以及瞭解使用者的狀態，並藉此提供適當的資訊服務，因此，情境感知（context-awareness）能力便成為重要的設計考量。當我們想利用人工智慧中的知識庫（knowledge base）來建構情境感知系統時；由於傳統的邏輯推理系統為單調邏輯（monotonic reasoning），但情境資訊（context）要透過裝設於環境中的感測器（sensor）來取得，變化的頻率十分快速。必須時常改變知識庫中的文句來確保情境資訊的有效性。但是任意修改知識庫勢必會違反邏輯單調性，於是如何有效管理知識庫中情境資訊的動態改變，便成為本論文的主要研究問題。 本論文提出一個以一階邏輯（first-order logic）為基礎的情境感知推理模型，將推理過程視為可以連續合理化執行的一連串動作（action），動作的執行將造成邏輯系統內部的狀態轉換，進而同步修改知識庫，讓系統能在單一個知識庫中管理情境資訊的增加與去除，系統則從目前的狀態中推理出下一步可執行的動作。進一步，我們利用規則（rule）來表示動作，並實做一個能支援情境感知推理的規則引擎來處理這些被定義好的規則。最後，我們利用該推理引擎開發三個不同應用環境下的情境感知應用系統，來測試其可行性。Context-awareness is one of the key requirements for intelligent systems in ubiquitous computing and smart space. A knowledge-based context-aware system infers the relevant services from its knowledge base, or KB, in which contexts are represented. To maintain relations between context KB and the physical environment synchronously, this thesis describes a logic-based model that attempts to management dynamic contexts stored in the KB from which a context-aware system can infer context-triggered actions and high-level contexts. By applying STRIPS representation and assumption, we model context-aware reasoning formally into a series of situation-driven actions. A context-aware rule engine called or CARE is developed based on the proposed reasoning model. To support making context-aware reasoning and building context-aware applications, CARE provides a systematic way to manage contexts in the KB. It utilizes a forward chaining rule engine called Jess to process rules, facts, and contexts. Three context-aware applications developed with CARE are described in this thesis to demonstrate the capabilities of context-aware reasoning, the design of context-aware applications, and the software support by CARE. All three applications are covered in healthcare domain. The first targets context-aware access control in pervasive healthcare environment. The second introduces a context-aware services integration system for assisting elder care in future smart home. The third aims to speedup the process of emergency visit by designing several location-aware notification services for medical members in an emergency department.Abstract...v List of Figures...x List of Tables...xii Chapter 1 Introduction and Motivation...1 1.1 Context-Aware Computing...2 1.2 An Elder Care Scenario at Smart Home...3 1.3 The View from Knowledge-Based Approach...5 1.4 Why Rule Base...6 1.5 The Research Challenge...9 1.6 Thesis Contribution...10 1.7 Thesis Organization...12 Chapter 2 Background and Related Work...13 2.1 Context-Aware Application...13 2.1.1 Context-Aware Application Framework...17 2.2 Knowledge-Based Context-Aware System...19 2.2.1 Ontology-Based Context Modeling...21 2.2.2 Context Management and Maintenance in the KB...25 2.3 Summary...26 Chapter 3 Context-Aware Reasoning...27 3.1 Reasoning with Context...28 3.1.1 The Context...28 3.1.2 Context Ontology...29 3.2 Situations and Actions...30 3.3 Context-Aware Reasoning System...32 3.4 Action Applications and Types...34 3.4.1 Sensory Actions...36 3.4.2 Primitive Predicates...38 3.5 Reactive Schemes...40 Chapter 4 Design and Implementation...43 4.1 Context-Aware Rule Engine...44 4.1.1 Action is Rule...45 4.1.2 Dealing with Context...47 4.1.3 Runtime Context Management...48 4.2 Software Design in CARE...51 4.2.1 Interface Design...52 4.2.2 Element Types of the KB...53 4.2.3 The Context Base...55 4.3 Event-Driven Design...55 4.3.1 Message Service Modules...57 Chapter 5 Building Applications with CARE...59 5.1 Context-Aware Access Control in Pervasive Healthcare...60 5.1.1 Context-Aware Authorization...61 5.1.2 System Architecture...65 5.1.3 Demonstration...69 5.2 Context-Aware Services for Elder Care in a Smart Environment...70 5.2.1 Context-Aware Elder Care...71 5.2.2 Context-Aware Services Integration System (CASIS)...73 5.2.3 Implementation...78 5.2.4 Demonstration...81 5.3 Context-Aware Notification for Emergency Care...86 5.3.1 Location-Aware Emergency Care System...87 5.3.2 System Architecture...92 5.4 Summary...95 Chapter 6 Conclusions and Future Work...97 6.1 Conclusions...97 6.2 Future Work...983698448 bytesapplication/pdfen-US情境感知知識庫邏輯推理規則推理引擎人工智慧context-awarenessknowledge baselogical reasoningSTRIPSrule enginethesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/53975/1/ntu-96-R93922070-1.pdf