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