An Efficient Tree-based Fuzzy Data Mining Approach

Chun-Wei Lin, Tzung-Pei Hong, and Wen-Hsiang Lu

Abstract

In the past, many algorithms were proposed for mining association rules, most of which were based on items with binary values. In this paper, a novel tree structure called the compressed fuzzy frequent pattern tree (CFFP tree) is designed to store the related information in the fuzzy mining process. A mining algorithm called the CFFP-growth mining algorithm is then proposed based on the tree structure to mine the fuzzy frequent itemsets. Each node in the tree has to keep the membership value of the contained item as well as the membership values of its super-itemsets in the path. The database scans can thus be greatly reduced with the help of the additional information. Experimental results also compare the performance of the proposed approach both in the execution time and the number of tree nodes at two different numbers of regions, respectively.


Keywords: fuzzy data mining, fuzzy set, quantitative value, CFFP trees, CFFP-growth, fuzzy frequent patterns.