We present a new method of modeling of cluster structure of a document collection and outline an approach to integrate additional knowledge we have about the document collection like prior categorization of some documents or user defined / deduced preferences in the process of personalized document map creation.
展开▼