Granules for association rules and decision support in the getRNIA system

Abstract

This paper proposes granules for association rules inDeterministic Information Systems (DISs) andNon-deterministic Information Systems (NISs). Granules for an association rule are defined for every implication, and give us a new methodology for knowledge discovery and decision support. We see that decision support based on a table under the condition P is to fix the decision Q by using the most proper association rule P\Rightarrow Q. We recently implemented a system getRNIA powered by granules for association rules. This paper describes how the getRNIA system deals with decision support under uncertainty, and shows some results of the experiment.