A PROPOSAL OF FUZZY MULTIDIMENSIONAL ASSOCIATION RULES

Abstract

Association rules that involve two or more dimensions or predicates can be referred as multidimensional association rules. Rather than searching for frequent itemsets (as is done in mining single-dimensional association rules), in multidimensional association rules, we search for frequent predicate sets. In general, there are two types of multidimensional association rules, namely interdimension association rules and hybrid-dimension association rules. Interdimension association rules are multidimensional association rules with no repeated predicates. This paper introduces a method for generating interdimension association rules. A more meaningful association rules can be provided by generalizing crisp value of attributes to be fuzzy value. To generate the multidimensional association rules implying fuzzy value, this paper introduces an alternative method for mining the rules by searching for the predicate sets.

Keywords

• Multidimensional Association rules
• Inter- dimension Association Rules
• Data Mining
• Fuzzy Sets.