Using clustering to filter results of an Information Retrieval system.