Granular Computing and Data Mining for Ordered Data: The Dominance-Based Rough Set Approach