UDDI(Universal Description Discovery & Integration)是一種跨平台的標準規範,用來描述並保存網路服務(Web services)的訊息,而人們可藉以發佈或發現某個網路服務的訊息。傳統的UDDI在實現描述與搜尋服務的方式上,一次只能在單一UDDI上利用商業服務實體的標識(keyword-base)及分類法(category)來進行查詢,對於網路服務發展迅速的時代,如此的模式已不能滿足使用者的搜尋需求,因此本論文的重點致力於改善目前UDDI各自為政的情況並提供一個統一的搜尋介面。 網域名稱系統(DNS)是多年前柏克萊大學發展出的一套階層式管理主機名稱對應IP的系統,透過階層式的架構,可以避免資料量過大的問題且維護工作進行較為輕鬆。 本研究將藉由網域名稱系統的樹狀結構及方便查詢之優勢,將UDDI從獨立提供服務改為階層式樹狀結構,進而讓使用者可以透過單一介面查詢到多台UDDI上的資料,不僅可讓網路服務更容易被查詢到,也讓使用者的查詢更為便利。
UDDI (Universal Description Discovery & Integration) is a cross-platform standard. It describes and stores information about web services. People can use it to publish or discover information regarding a web service. The old-fashioned implementation of UDDI only allows you to search one UDDI using business services entity’s label (keyword-base) and classification methods (category). This simplistic search methodology simply is not enough for today’s users. Here we try to solve this problem by creating a unified search UDDI interface. Domain Name System (DNS) is a hierarchical hostname to IP lookup system, which was developed so many years ago by Berkeley University. It manages classified IP addresses corresponding to the Fully Qualified Domain Name (FQDN) through the hierarchical structure for translating meaningful domain names to humans. By using a hierarchical framework, we can distribute workload and simplify maintenance. We will create a UDDI system using a hierarchical tree-like structure. The UDDI will migrate from individual servers to a hierarchical tree-like structure and allow users to meta-search a number of UDDIs. Not only web services will be easier to discovery, searching itself will be even more convenient.