A Dual-Reduct Approach to Generate Core Rules

Abstract

In this paper, we propose a Dual-Reduct method to generate core rules from original data sets for decision making. We rank the rules by rule usefulness after the step of first reduct. Then we take the useful rules as condition attribute and construct another new decision table. After the step of second reduct we generate core rules from the new constructed decision table. In our approach the generation process is straightforward and objective. At the same time, our approach can significantly reduce the number of rules comparing to the traditional generation approach because we adopt rule usefulness as a measure of core rules. We also provide theoretical proofs and deductions. Our approach is proved to be feasible and effective in a production security system.