Ranking Weak-Linked Documents on the Web

Abstract

The current commercial Web search engines do a good job at ranking web pages with hyperlink information. However, there are also many common documents such as PowerPoint files or Flash files which do not have enough hyperlink information. We call such documents weak-linked documents. Current search engines return therefore either completely irrelevant results or poorly ranked documents when searching for these files. This paper addresses this problem and proposes a solution: RoC (Ranking weak linked documents based on Clustering). For a given query q, RoC (1) first clusters traditional Web page search results in order to find what topics existing on the WWW are interesting to the users, (2) then assigns a weight to each topic cluster based on the ranks of the web pages in it, and finally (3) ranks all relevant weak-linked documents based on their similarity to the weighted clusters obtained from the Web. The experiments show that our approach considerably improves the result quality of current search engines and that of latent semantic indexing.