多維度情境感知資訊擷取與交換之研究

Abstract

「情境感知式電腦運算」的概念是早先在「普遍運算」研究方面所引入的概念之一。並這些概念從那時就受限於既存在的其他研究。「情境」是談到電腦運算裝置所嵌入的實體或社會的處境跟環境。情境感知運算是要取得和運用與一架機器有關的這種情境，以提供適合特定環境的服務。在過去許多的情境感知系統都只使用單一情境感知技術去反應週遭的環境變化與使用者資訊，往往無法完整的描述目前環境之變化，系統給予使用者的資訊回饋也會有所侷限。 有鑒於此，本研究透過建立多維度廣泛用途的情境感知資料格式加上後端管理系統的映射機制搭配資訊處理，該廣泛用途的情境資料格式包含了情境資訊中的“人、事、時、地、物”等五種情境資訊，讓資訊的流通可以更自動化、彈性化、個別化的將情境資訊呈現給每個使用者。另外，本研究依據不同類型使用者情境，建立了「人」與「物」兩種不同類型的使用者情境應用，情境感知系統會依據使用者的類別透過情境資訊的映射機制來提供不同的情境感知環境給使用者，再加上服務可以維持長時間或短時間的提供使用者服務，可簡化需要使用者頻繁的啟動使用者應用端程式，以增加使用者的便利性。

Context-aware applications are getting popular in modern life. Various kinds of context data are used to reflex the situations and events when users interact with the computer applications. Many context-aware applications are developed in recent years. However, most of them only use single context data to represent the scenario that users might confront in the applications. In this research, we develop a multi-dimension context-aware operation model which can be used in most of the context-aware applications. A general purpose context-aware message exchange mechanism is proposed in our model. Five types of context-aware information including : “person”, “event”, “time”, “area”, and “object” are used to reflect the situations and events in an application. In addition, a prototype which combines various context data such as RFID, GPS and time is implemented to provide an integration of context-aware applications such as “person” and “goods”. In order to provide a flexible manipulation of back-end services, we define a mapping mechanism which can provide a configurable mapping among context data and services. Using our model, a developer can establish a new context-aware applications and services very easily.