以網頁為基礎的智慧型電子化服務櫃檯之設計與實作

Design and Implementation of a Web-Based Intelligent e-Service Counter

楊豐兆(Feng-chao Yang)

《大葉學報》 12卷2期 (2003/12) Pp. 9-18

https://doi.org/10.7119/JDYU.200312.0009

摘要

本論文目的在提出一個以網頁為基礎的智慧型電子化服務櫃檯（intelligent e-service counter），並命名為Dr. Web，此一整合性網路服務系統包含以電子化服務為基礎的資料提供、資料分析與櫃檯等三個模組。資料提供模組除了存放有關客戶的歷史交易與客戶基本資料，為有效建立商業領堿知識庫，透過交談式的使用者介面可以協助使用者建立領堿相關知識庫。資料分析模組使用商業知識庫中的商業規則對客戶的歷史性資料進行推論與分析，其結果交由櫃模組直接呈現在客戶端網頁。為使電子化服務元件的開發過程能模組化、電子化服務元件的使用更具彈性，本系統使用UML（unified modeling language）分別對這三個模組進行分析與設計，最後整合為電子化客戶關係管理應系統。這些模組透過分散式合作解題計畫（cooperative distribute problem solving plan）能完成模組與模組間、以及各類型服務櫃檯與服務櫃檯間的協調與合作。在本文中舉例之系統是一個允許服務提供廠商穨過後端介面輸入專業領堿的背景知識（background knowledge），或是轉換已解答之常問問題（FAQ）並將之儲存在知識庫中，客戶可以直接利用網頁上的服務櫃檯直接提出問題或進行搜尋，經由我們所設計的推論機分析客戶過去的消費行為特性與企業本身的產品特性後，以主動或被動的方式提供客戶適切的服務。經數個實際範例證明本系統的使用，能在良好的互動關係下對客戶進行充份瞭解，確實可以達成預期的目標。

關鍵字

電子化服務櫃檯 ； 客戶關係管理 ； 分散式合作解題計畫 ； 知識庫 ； 推論機

並列摘要

This paper presents a web-based intelligent e-service counter, named Dr. Web. This system includes a data-providing module, a data-analysis module and a counter module to achieve an integrated Internet e-service. Besides storing the personal data and the trade-history records of customers, the data-providing module is capable of assisting users in effectively establishing the business-domain knowledge base through a user-friendly interactive interface. As for the data-analysis module, it utilizes the business rule formulated in the business-domain knowledge base to infer and analysis past customer records. The counter module will finally receive results from the data-analysis module and display them on the client web page directly. In order to modularize the development procedures of e-service components and make them flexible when using the e-service components, the Unified modeling Language (UML) is used to respectively analyze and design the three modules which are integrated as an application system of electronic customer relationship management. These modules can achieve the coordination and cooperation among modules and service counters through the cooperative distributed problem-solving plan. An example system illustrated in this paper allows the service-providing companies to input the background knowledge of the professional domain behind the interface, or to transfer the solved FAQ into the knowledge base. Customers can ask questions or search for information directly at the service counter on the web page. Based on the past consumption behavior of customers and the product properties of enterprise, the well-designed inference engine will respond and provide an appropriate service for the customers actively or passively. Form the results of several examples, it is proved that Dr. Web is a good system in understanding the customers based on good interaction relationship, and satisfies the expected purpose.

並列關鍵字

e-service counter ； customer relationship management ； cooperative distributed problem solving plan ； knowledge base ； inference engine