服務導向架構(Service Oriented Architecture,SOA)的概念已經愈來愈受到研究的重視,企業也逐漸的採用服務導向的思維來看待整個公司的整體,而網路服務(Web Service)應用系統是很適合用來實踐服務導向架構的一個軟體架構。企業本身營運所需的業務,透過服務導向的的概念,建置各式各樣本身所能提供的服務。由於企業營運的過程中,需要大量的服務來支援,隨著企業的成長網路服務的數量只會愈來越多,當這些服務大量的被建立出來之後,過多的服務會造成在服務管理以及選取上的負擔。 為了管理產生的這些大量服務,並且要以一個彈性的系統架構來讓這些服務在上面運作,在本研究中,我們剖析網路服務定義文件,擷取服務的名稱及輸出輸入資訊,進行編碼轉化為服務向量矩陣,透過自適應共振理論(Adaptive Resonance Theory)的學習機制,以物以類聚的特性,自動為系統中的服務做分群,產生數個服務聚類,協助服務的管理;並透過一個服務池(Service Pool)的概念,輔以入口網站的思維,以推薦系統(Recommendation System)的形式,建立一個彈性的架構,方便企業管理及使用營運所需的服務。
The conception of Service Oriented Architecture (SOA) becomes more and more important not only in research field. The industry field also gradually adapts SOA as its thought to apply on the companies. The Web Service Application System is a suitable software for the one who wants to adapt SOA framework. Using the concept of SOA, the enterprises construct their operation business into various providerable services. With the growth of enterprise, there are more and more demands on the internet service, which leads into a burden on service operation and selection. We try to use an elastic system framework to manage this problem. In this research, we analyze the definition documents of internet service; try to get the service names and the information of input and output; transfer the code into the matrix; utilize the learning behavior of the Adaptive Resonance Theory to classify all services into groups to assist doing management. We also adapt the concept of Service pool with the idea of portal site to construct an elastic framework for the business management and operation in the way of Recommendation System.