on simultaneous clustering and cleaning over dirty data published presentations and documents on DocSlides.