skip to main content

情境感知系統中使用XML綱要映對的一般化解譯器

何祥意; He, Shiang-yi 錢炳全; Been-Chian Chien; 資訊工程學系碩士班 2009

線上取得

  • 題名:
    情境感知系統中使用XML綱要映對的一般化解譯器
  • 著者: 何祥意; He, Shiang-yi
  • 錢炳全; Been-Chian Chien; 資訊工程學系碩士班
  • 主題: 情境感知; 情境解譯器; 綱要映對; 綱要配對; Schema matching; Context-awareness; Context interpreter; Schema mapping
  • 描述: 早期的情境感知系統在系統架構上是採用緊密結合(Tightly coupled)的設計方式,亦即針對單一特定的系統需求、應用環境與服務對象,將系統的各個層次,從感知器、驅動程式、情境定義到執行系統等,均特別訂製、完全結合並相依。這樣的系統構建方式雖然在效能與系統完整性上有很好的表現,然而當需求、服務對象或應用環境有所變動時,此一系統便無法提供正常的服務,並且也不易維護;另一個問題則是系統的可攜性太低,不利於一般情境感知應用系統的建構與普及。本研究主要針對以上的問題,將情境資訊(Context)的解譯過程視為感測資料與情境資訊的映對(Mapping)關係提出以解譯描述語言(Interpretation script)為主的一般化情境解譯器架構,發展圖形化的情境映對編輯工具(Context mapping editor),並且設計綱要配對(Schema matching)演算法以用自動映對並產生情境解譯描述語言替代傳統的解譯程式碼,用以將感測資料解譯為情境資訊。在系統維護及設備更新方面,本論文所提出的方法也相較於以往需直接修改解譯器內部程式碼的方式更為方便,並具彈性,建構時間上也相對地節省。實驗顯示,本論文所提出的SYM演算法在綱要的配對及綱要的檢索方面皆有良好的效能,能以快速又精確的方式套用合適的歷史映對資料,使所提出的一般化解譯器架構能夠達成快速產生解譯器之目標。
    Traditional context-ware systems were tightly coupled and a system was designed for specific requirements, service applications, environments, and persons. Such a system is always implemented by linking different layers including sensors, drivers, context information and system execution tightly. Although this constructing approach is good in system performance and functionality, systems will fail while the requirements, service or environments are changed. Furthermore, it is difficult for users to maintain the system. The low portability is another serious problem since it is hard to reproduce and construct another context-aware application fast. In this thesis, we regard the process of context interpretation as a mapping from sensor data to context and a generic architecture for context interpreter based on interpretation script is proposed. We develop a graphical context mapping editor and a schema matching algorithm to map and generate interpretation scripts automatically for replacing program code of interpretation. With the support of interpretation scripts, the generic interpreter can be able to interpret each sensing data into context. The proposed approach is more convenient and flexible than the traditional approach which has to modify internal code of interpreter directly. It is also good for system maintenance and reducing construction time. The experimental results show that the proposed SYM algorithm has a good performance in schema matching and historical data searching, which can provide a fast and accurate method to select appropriated historical mapping schema and accomplish the objective of generating context interpreter.
    碩士
  • 建立日期: 2009
  • 格式: 121 bytes
    text/html
  • 語言: 中文
  • 識別號: http://nutnr.lib.nutn.edu.tw/handle/987654321/2859
  • 資源來源: NUTN IR

正在檢索遠程資料庫,請稍等