- 期刊
應用網站使用探勘於網站瀏覽之個人化－以健康檢查業為例
Applying Web Usage Mining to Personalization of Web Browsing-A Case Study in the Health Examination Industry
摘要
在網路盛行的今日，人們經常利用網際網路找尋所需的資訊。由於不同的使用者會有不同的需求，使得網站瀏覽之個人化已成為網站經營的必然趨勢，網站将針對個人來設計特定的內容及相關的資訊服務，與顧客建立起一對一的關係。 在網站上的日誌檔有使用者瀏覽的記錄，隱含著使用者的需求資訊。本研究選擇醫療產業中之健康檢查業為研究對象，利用資料探勘技術於網站日誌，以挖掘出使用者存取網頁的型樣。網頁之瀏覽次數固然重要，但網頁停留時間（即網站黏性）才眞正說明網頁上的內容被用戶所關注的程度。因此，為了達到網站瀏覽之個人化，本文提出以網頁停留時間為支持度權重之關聯法則來探勘網頁之間的關聯，我們稱之為WS演算法。另外，本文亦将網站日誌與會員基本資料結合，利用資料探勘軟體之決策樹功能來針對使用者特性與瀏覽行為進行分類。網站瀏覽之個人化，不僅可以提供使用者快捷且符合需求的資訊，並且可以幫助健檢業者改善網站設計架構及推出有效的廣告行銷。
並列摘要
Nowadays, in the prosperous era of Internet, people usually use the Internet for accessing the information they need. Because different users may have different needs, personalization of web browsing has become the trend for running a web site. A web site will be designed with the specific contents and relevant information that meet individual user's need and will establish a one-to-one relationship with customers. Logs on a web site keep track of browsing records of the user and conceal the user's demand on information. The subject of this study is on health examination in the medical care industry. By utilizing data mining techniques on web logs, we can find out the pattern in which users access web pages. Although the number of accessing a web page is an important consideration, the duration in which a person browses the web page really reveals the extent a user concerns about the web page. Therefore, in order to achieve personalization of web browsing, we propose an association rule algorithm to discover the association between web pages. This algorithm uses duration on web pages as the weighted support and is called the WS algorithm. In addition, by combining web logs and membership data, we utilize the decision-tree function of data mining software to classify the characteristics and browsing behavior of users. Personalization of web browsing not only can provide the user with fast access to the required information, but also can help the health examiner to improve the design of the structure of web pages and propose an effective advertisement.