Document Clustering in Personal Dataspace

Abstract

In Personal Dataspace (PDS), documents containing a lot of useful information play an important role in our daily work. However, it is difficult to manage the information in these documents efficiently. In this paper, we first extract some frequent terms from documents, and then cluster these documents based on the terms. Thus users can query the documents based on their contents conveniently. The experiments demonstrate the accuracy and efficiency of the key techniques in our approach.