Meta-Rule-Guided Mining of Association Rules in Relational Databases.

A metarule guided data mining approach is proposed and studied which applies metarules as a guidance at nding multiple-level association rules in large relational databases and is powerful and eecient in data mining from large databases.

Abstract

A meta-rule-guided data mining approach is proposed and studied which apphes meta-rules as a guidance at finding multiple-level association rules in large relational databases. A meta-rule is a rule template in the form of "P A ... A P, --* Q A...A Q,,", in which some of the predicates (and/or their variables) in the antecedent and/or consequent of the meta-rule could be instantJared. The rule template is used to describe what forms of rules are expected to be found from the database, and such a rule template is used as a guidance or constraint in the data mining process. Note that the predicate variables in a meta-rule can be instantJared against a database schema, whereas the variables or some high-level constants inside a predicate can be bound to multiple (but more specific) levels of concepts in the corresponding con- ceptual hierarchies. The concrete rules at different concept levels are discovered by a progressive deepening data mining technique similar to that developed in our study of mining multiple-level association rules. Two algorithms are developed along this hne and a performance study is conducted to compare their relative efficiencies. Our experimental and performance studies demonstrate that the method is powerful and efficient in data mining from large databases.