Characterization and evaluation of similarity measures for pairs of clusterings