A binary method for fast computation of inter and intra cluster similarities for combining multiple clusterings