以本體論為基礎具推論能力之可隨插式服務平台-以家庭照護為例

Abstract

傳統的應用程式在建構時主要以是服務提供者的觀點來設計，但這往往忽略了服務使用者真正的需求。就服務使用者的立場而言，使用者真正需要的服務可能是來自於不同的服務提供者，並且這些異質性的服務通常是無法溝通合作的。此時使用者便需要一個平台來整合異質性的服務。針對這個問題，我們提出了一個以本體論為基礎的可隨插式服務平台，希望能夠在居家照護的環境下依據服務使用者的需求整合異質性的服務，並且以健康忠告者(Health Advisor)為例子。 在本研究中我們運用本體描述語言(OWL)以及語意網法則語言(SWRL)來定義可隨插式本體論服務。本體描述語言是W3C所定義的語意網標準語言，而語意網法則語言是延伸至本體描述語言的一種法則語言。根據本體論內服務的描述，代理人(agent)可以推衍出高階的資訊、偵測環境的變化、進行個人化的推薦、以及與系統中其他的代理人溝通合作。透過可隨插式本體論驅動的技術，我們可以建構出更具有彈性、可調適性、且能夠自我管理的居家照護應用系統。

Traditional applications are designed based on service providers’ viewpoint, not users’. And it is a difficult task to customize applications for each user. On the other hand, service demanders need a platform to integrate individual services supplied by different providers. The cooperation between these heterogeneous services is a key issue we need to consider. To address this problem, we propose an ontology based pluggable service platform to seamlessly integrate heterogeneous services which are exact what user wants in home care environments, and take Health Advisor as our example. In this paper we use Web Ontology Language (OWL) and Semantic Web Rule Language (SWRL) to specify a pluggable ontology service. OWL is W3C standard language for Semantic Web, and SWRL is a rule language developed by the DARPA Agent Markup Language (DAML) Program, extending directly from OWL Semantics. By plugging different ontology service, the agents in our system can infer high-level context information, detect events, give recommendations and cooperate with other agents. Through the use of ontology-driven development and reuse techniques, we can achieve a more flexible, adaptable, cooperative, and self-managing home care applications.