BLOGRANK: Ranking on the blogosphere

Apostolos Kritikopoulos, Martha Sideri and Iraklis Varlamis

Although, the Blogosphere is part of the World Wide Web, weblogs present several features that differentiate them from traditional websites: the number of different editors, the multitude of topics, the connectivity among weblogs and bloggers, the update rate, and the importance of time in rating are some of them. Traditional search engines perform poorly on blogs since they do not cover these aspects. We propose an extension of Pagerank, which analyses and extends the link graph, in an attempt to exploit some of the weblog features. The analysis of the weblogs’ link graph is based on the assumption that the visitor of a weblog tends to visit relevant or affiliated weblogs. Our algorithm, BlogRank, models the similarity between weblogs by incorporating information on common users, links and topics and generates a global ranking for all weblogs in a set. To validate our method we ran experiments on a weblog dataset, processed and adapted to our search engine: http://spiderwave.aueb.gr/blogwave Our experiments suggest that our algorithm enhances the quality of returned results.

Poster

Available as PDF