隨著網際網路的普及,愈來愈多人將網路作為一項資訊來源的主要管道,透過關鍵字搜尋來取得資訊。由於有關關鍵字搜尋行為的研究至今為止仍然不多,故本研究欲探討網友的關鍵字搜尋行為。本研究利用創市際的線上資料庫,收集1338位使用者兩個月內的Yahoo!關鍵字搜尋記錄進行資料探勘,從中找出哪些網路行為與人口統計變數對於關鍵字的搜尋行為有顯著之影響。並進一步將每一個關鍵字區分為熱門或冷門關鍵字,再找出此兩類關鍵字使用者在網路行為與人口統計變項分佈上的差異所在。 研究結果發現,(1)年齡、職業與收入對搜尋關鍵字數量之影響顯著不同;(2)搜尋寬度、長度及搜尋關鍵字數量有顯著影響;(3)搜尋熱門字的使用者與冷熱門搜尋者在搜尋數、寬度及長度上均有所差異,但在瀏覽深度、網路停留時間與搜尋引擎使用率上則無差異。
As the Internet surges, more people deem the Internet as a main information source, and usually use keyword search to acquire information. Since the researches about search behavior were still very few, this study intends to explore online users’ keywords using and searching behavior. The study collected two-months Yahoo search clickstream data from InsightXplorer’s ARO database. First, we tried to how users’ demographic traits to affect their keyword searching behavior. Next, we divides all keywords into two types(hot words and cold words), and find the behavioral differences between the different keyword search approaches. The result showed the following findings: (1) User’ age, occupation and income have significantly different impacts on keyword search volume. (2) Users’ browsing width and length will significantly increase their keyword search volume. (3) Several differences of information behavior and demographic traits(gender, age, occupation, education, and income) exist between user using different approaches to search keyword(only hot words vs. all words).