Summary: | 碩士 === 大同大學 === 資訊工程研究所碩士班 === 92 === In the recent years, the development of the Internet is so quickly and the ways of using Internet are very easy and convenient, so the population of surfing WWW increases fast. Because of huge information on the Internet, user may give up to find the information that he really wants after a lot of browsing behavior. In this thesis, we discuss how to analyze and clean out the history browsing behavior of the website. And then, find out the contents of the user''s surfing habit or user''s interested web pages. When the user browses this website again, the web page recommendation mechanism will recommend the user the web pages that user is interested. This thesis employs the sequential association rule, combines the genetic algorithms and back-propagation network (BPN) to construct a system with ability of predicting and offers individualized basis that the web pages are recommended from the website.
|