A Survey on Improving the Web Search Ranking by User Behavior Information

Abstract

In this paper we propose new methods for ordering the Web pages returned from search engines. Given a few search keywords, nowadays most search engines could retrieve more than a few thousand Web pages. The problem is how to order the retrieved Web pages and then to present the most relevant Web pages first. We propose new factors to allow relevant Web pages to be ranked higher. The factors include keyword popularity, keyword to Web page popularity, and Web page popularity. The keyword to Web page popularity records which Web pages have been selected corresponding to the search keywords. The Web page popularity determines how often the Web pages have been selected and also how many popular keywords are contained in the pages. Using these popularity factors, our system is able to rank more popular pages higher, which will help most search engine users find the more popular and plausibly the more relevant pages.