以可能性派翠網路為基底之服務配對; Possibilistic Petri-Nets as a Basis for Service Matchmaking

NCU Institutional Repository > 資訊電機學院 > 資訊工程研究所 > 博碩士論文 > Item 987654321/8500

請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/8500

題名:	以可能性派翠網路為基底之服務配對;Possibilistic Petri-Nets as a Basis for Service Matchmaking
作者:	王躍強;Yao-Chiang Wang
貢獻者:	資訊工程研究所
關鍵詞:	服務導向多重代理人;服務註冊機;服務配對;可能性派翠網路;possibilistic Petri-nets;multi-agent system;service oriented architecture;UDDI;service matchmaking
日期:	2006-10-16
上傳時間:	2009-09-22 11:28:37 (UTC+8)
出版者:	國立中央大學圖書館
摘要:	近年來，軟體即服務的概念逐漸在軟體開發中被注意以及採納，著名的軟體元件平台如多重代理人、網際服務等技術的成熟，使得軟體元件的可重複利用性在日益複雜的系統中日漸展現其優勢，然而在一個分散且開放的軟體平台中，如何搜尋適當且符合需求的服務就變成是一個重要的關鍵，這包含軟體開發階段的靜態搜尋與軟體執行階段的動態搜尋等方案。要達成這樣的目標，需要一個服務配對的平台，該平台的建置將探討以下相關的議題，包含如何描述與註冊一個軟體服務、如何描述一個服務需求、以及如何有效的配對服務需求與註冊的服務等等。本研究提出一個以可能性派翠網路為基底的服務配對機制，此機制透過可能性推論與可能性派翠網路表達不確定的特性，來處理服務配對中，因為需求與服務之間的不一致性所造成的語意以及部份相似的問題。並進一步透過可能性派翠網路代理人描述語言來視覺化模組整個服務配對的過程，其中包含服務需求與服務提供的表達，兩者之間的配對關聯，以及服務配對的運行和配對結果。同時，為了方便描述與使用，我們利用XML 與Ontology 等技術，來發展結構化的服務描述規格，與表達服務領域相關資訊，以提昇服務配對的精準度。此外，針對網際服務架構與網際服務語言方面，也提出與可能性派翠網路配對機制結合的方式，以整合標準的UDDI 服務註冊機，並透過自訂的技術標籤(tModel)來擴充與提升 UDDI 註冊機的搜尋速度與精準度。並藉此提出新型態的UDDI 瀏覽器，透過服務描述語言的抽象資訊，方便網際服務的開發者以更有效率與方便性的方式，瀏覽所需要的網際服務。最後，我們提出一個以服務為導向的多重代理人系統，來驗證我們所提出的方法，並透過中間代理人為橋接器，來整合不同平台上的服務資源。歸納上述所言，本研究包含以下的特點與貢獻： ⣶98; -以可能性推論與可能性派翠網路模型為基底之服務配對平台。 ⣶98; -結合XML 與Ontology 技術於結構化服務描述語言與服務配對機制的使用，達成語意配對與部分配對的功能。 ⣶98; -整合代理人與網際服務平台上的服務配對。 ⣶98; -與現有的UDDI 標準結合，並擴充其搜尋能力與提升搜尋效能。 ⣶98; -提出服務為導向的多重代理人系統來整合異質的服務資源。 Recently, the concept “software as a service” has attracted the researchers’ attention and adopted by the software development communities gradually. Multi-agents and Web services are well-known software component technologies that are mature and can be lever- aged to demonstrate the importance of reusability of software component in developing a complex application system. However, how to locate a suitable service component becomes a crucial issue in an open and distributed application environment. Statically matching service in software development phase and dynamically matching service in software exe- cution phase are approaches covered by this issue. To achieve such goal, we need a service matching platform which deals with topics, including how to describe and register ser- vices, how to represent requests, and how to efficiently match services with requests. This research proposes a possibilistic Petri-nets (PPN) based service matchmaking mechanism which leverages possibilistic reasoning and PPN formalism to handle the uncertainty fea- ture caused by the inconsistency between requests and services when performing service matchmaking. Additionally, a specification called PPN-ASDL (PPN-based Agent Service Description Language) is devised to visually represent the requests, services, the service matchmaking between them, and what the matching results would be. The XML and On- tology technologies are involved to develop structural language specification and to capture domain information for ease of usage and helping the matching process by increasing the precision of results. On the other hand, we propose an approach to incorporate WSDL specification into the PPN-ASDL matchmaking model and integrate it with the standard UDDI registry service with newly-defined tModels to extend the functionalities and perfor- mance of UDDI registry service. Meanwhile, a new type of UDDI browser that is capable of directly browse Web service information through meta-data of WSDL helps the developers deal with Web services more efficiently and easily. Finally, we proposes a system called SAM (Service-oriented Architecture for Multi-agents) to reach the wide adoption and rou- tine use of agents, Web services, and semantic web technologies through the middle agent (a PPN-based service matchmaker). Following features and contributions are summarized as the conclusions of this research: • elaborate a service matchmaking platform by leveraging possibilistic reasoning and possibilistic Petri-nets. • incorporate XML and Ontology technologies in developing structural service descrip- tion language and capturing domain knowledge to cope with sematic match and partial match. • merge service matchmaking for multi-agents and Web services environments. • integrate and extend UDDI standard with newly-devised searching and browsing func- tionalities. • propose a service-oriented architecture for multi-agents to integrate heterogenous ser- vice resources.
顯示於類別:	[資訊工程研究所] 博碩士論文

文件中的檔案:

檔案	大小	格式	瀏覽次數

在NCUIR中所有的資料項目都受到原著作權保護.

社群 sharing

資料載入中.....