Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
379230 | Data & Knowledge Engineering | 2008 | 14 Pages |
Abstract
Measuring association among variables is an important step for finding solutions to many data mining problems. An existing metric might not be effective to serve as a measure of association among a set of items in a database. In this paper, we propose two measures of association, A1 and A2. We introduce the notion of associative itemset in a database. We express the proposed measures in terms of supports of itemsets. In addition, we provide theoretical foundations of our work. We present experimental results on both real and synthetic databases to show the effectiveness of A2.
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Animesh Adhikari, P.R. Rao,