知識本體與推理於情境感知訪客指引之研究

Abstract

在許多公共設施與遊客中心，訪客協助機制能夠提供給每位訪客有用的 資訊。然而，這些資訊多為靜態且流於泛泛，因此需要有服務人員能依據訪客的位 置、興趣與偏好，提供個人化的協助。例如，當新生在參訪大學校園 時，會想要知道符合自己研究興趣的教授有空的時間，而這些資訊是可以藉由現在的情境推論而知的。 本研究提供一種知識表達法與推論流程，以提供訪客情境感知的協助 。我們發展了一個用以表達時間、室內地點、事件、路線 、個人資料與研究主題等概念的OWL知識本體，以及26條用來推論訪客協助的SWRL規則。從實驗結果得知，此一情境感知的推論方法，能夠用以尋找訪客 的位置、目的地、無法使用的通道、有空閒的教授、適當的會面場所 ，以及與訪客研究興趣相符的教授等。

In many public facilities and tourist destinations, visitor assistance provides useful information to all visitors. While much information is generic and static, a human assistant can offer more personalized assistance by reasoning about the visitor’s location, interests and preferences. For example, a new student visiting a university campus may need information on finding the professors with matching research interests at their available times, which may be derived by reasoning about the current contexts.

This research explores the knowledge representation and reasoning process required to provide context-aware assistance to the visitors. In this thesis, we have developed the OWL ontology for time, indoor location and event context, as well as for routes, personal profiles and research topics. We have constructed 26 rules in SWRL to infer context-aware visitor assistance. A sample experiment has been conducted to demonstrate our context-aware reasoning process. The proposed approach can be used to find the visitor location, destination, unavailable passage, available professors and suitable location or person based on the visitor’s research interests.