Slovak Blog Clustering Enhanced by Mining the Web Comments

Abstract

Web environment gives us an opportunity to take broader aspects of textual information into account, for example information about behavior of web users, global web trends or social interactions between users. In this paper we present and evaluate method suitable for enhancing blog clustering using the information hidden in web comments. We found out that blog clusters based on clustering of the commentators differ significantly from the content clusters. But according to the results of our experiments implicit relations between commentators can be used in addition to content clustering and improve the quality of content clusters.