A Fuzzy, Incremental, Hierachical Approach of Clustering Huge Collections of Web Documents

Abstract — Since every day millions of posts are published inside the blogosphere a huge collection of web documents develops. Clustering this ever-changing collection is a very time consuming task. Therefore some certain challenges has to be accomplished because a clustering cannot be executed from scratch all the time. The presented fuzzy, incremental and hierarchical clustering algorithm tries to succeed these challenges with both, clustering terms and documents in a meaningful way and keep them up-to-date all the time. Furthermore we take a critical look at the performance, which is crucial on such a live data collection