| Summary: | 碩士 === 中國文化大學 === 資訊管理研究所碩士在職專班 === 93 === Face to the network there is such a great deal of in formation. Users not only want the speed of searching in internet more quickly, the data searching range to be wider, but also find the data that users themselves desired in most short time. While the user browsing websites, We find that the most of websites not to ask the user key in their ID, or they are not member websites. Web log recorded every user from come the website to go to leave. The web page browsed and every kind of behavior that carry on. Such huge data and that increase rapidly. Implicit the real needs for user.
So this article announces the mechanism of web page recommends. Through ac-cumulate the record of browsing, Then analyzes the interests of the each user. While the next user browsing web page, It will browse the well timed recommendation suits of the web page. Using the data mining in web log record, In order to find out all the kinds of the web page for user accesses.
Generally speaking, Only excavates the web page is browsed or not, Did not bring time into the consideration, We can't really know how much the favorite degree is to the user in that web page. So this research includes stay time. Making use fuzzy as-sociation rules to analysis, In order to find the relativity between the web pages. Creat-ing a series recommends of the web page. Every different user could get the well timed recommendation to the next web page or books that browses to in the website, To de-crease browsing time and find the information we are looking for more quickly.
|
|---|
