An Improvement of Fuzzy Association Rules Mining Algorithm Based on Redundancy of Rules

Abstract

In data mining approach, quantitative attributes should be appropriately dealt with as well as Boolean attributes. This paper presents an essential improvement for extracting fuzzy association rules from a database. The objective of this paper is to improve the computational time of mining and to prune extracted redundant rules simultaneously for an actual data mining application. In this paper, we define the redundancy of fuzzy association rules as a new concept for mining and prove essential theorems concerning the redundancy of fuzzy association rules. Then, we propose a basic algorithm based on the Apriori algorithm for rule extraction utilizing the redundancy of the extracted rules. The essential performance of the algorithmis evaluated through numerical experiments using benchmark data. Fromthe results, themethod is found to be promising in terms of computational time and redundant-rule pruning.