Interactive Analysis of Preference-Ordered Data Using Dominance-Based Rough Set Approach

We present a method of interactive analysis of preference ordered data that is based on Dominance-based Rough Set Approach (DR.SA). The presented here methodology is conceptually similar to multi-dimensional reports (pivot tables) applied in On-Line Analytical Processing (OLAP). However, it allows to identify patterns in data that remain undiscovered by traditional approaches to multi-dimensional reporting. The main difference consists in use of specific dimensions and measures defined within DRSA. The method permits to find a set of reports that ensures specified properties of analyzed data and is optimal with respect to a given criterion. An example of reports generated for a well-known breast cancer data set is included.