運用資料採礦於台灣觀光資訊入口網站瀏覽行爲分析

A Study of Applying Data Mining Technology to Browsing Behavior Related to the Tourism Information Portal of Taiwan

王佳鳳 ； 黃正聰 ； 郭柏村

《商業現代化學刊》 4卷3期 (2008/03) Pp. 55-66

https://doi.org/10.6132/JCM.2008.4.3.05

摘要

台灣觀光資訊入口網站(http://taiwan.net.tw)每天約有一萬人次上網，在這無形中已經儲存著數以兆計由0與1所組成的磁性記憶與資料，我們運用資料採礦技術嘗試將龐大資料轉化成寶貴的資訊以供相關決策人員參考。
本研究時間爲2004/2/1到2004/4/30共三個月時間，針對台灣觀光資訊入口網站之網站日誌檔(log files)透過WEB TREND、SPSS、SPSS CLEMENTINE等軟體工具，對其作相關瀏覽行爲之分析。
主要分析結果爲：一、網路使用方面：每人上網瀏覽頁數平均爲四頁，且停留時間以一分鐘內占最多人斂，上網時段以平日人數較多於週末，一天內平均有三階段上網人潮，分別爲早上九點到十一點、下午三點到五點、晚上八點到十一點。二、交叉分析方面：點選時間與旅遊型態有顯著差異。三、關聯分析方面：不同旅遊景點類型其相關點選的旅遊景點不同。本研究最後並提出相關建議及策略供產、官、學界參考。

關鍵字

資料採礦 ； 關聯分析 ； 網站日誌檔 ； 台灣觀光資訊入口網站

並列摘要

There have been at least 10, 000 people browsing the Tourism Information Portal of Taiwan (http://taiwan. net. tw/). And, billions of bytes of related magnetic memory and data have already being stored. These are rated ”0” to ”1”. Large amounts of data are transferred into valuable information, via a technique of Web Mining, which provides information for decision-makers.
The study period for this research was three months-from 2/1/2004 to 4/30/2004-focusing upon the log tiles of the Tourism Information Portal of Taiwan. The data was, then, analyzed via WEB TREND, SPSS, and SPSS Clementine.
The results were as follows:
1. Internet usage: On average, each person browses four pages each day, and most people's browsing time is within one minute. There are more people surfing the Internet during weekdays, and there are fewer people surfing the Internet on weekends. There are three major, definable periods of Internet usage from 9:00 am to 11:00 am, from 3:00 pm to 5:00 pm, and from 8: 00 pm to 11:00 pm.
2. Cross-analysis: The ”click time” and ”types of traveling” show great differences, among users.
3. Connected analysis: The different types of tourist attractions visited at the site are quite different from the click time for these tourist attractions.
We offer suggestions and strategies, according the above results, to industry executives, governmental officials, and academics.

並列關鍵字

Web mining ； Relational Analysis ； Web site log tiles ； Tourism Information Portal of Taiwan