Mining quantitative association rules under inequality constraints

In the past several years, there has been much active work in developing algorithms for mining association rules. However, in many real-life situations, not all association rules are of interest to the user. A user may want to find association rules which satisfy a given inequality constraint for a set of quantitative items. In other words, users are more interested in the subsets of those associations. We present how to integrate the inequality constraints into the mining process and reduce the number of database scannings. The algorithm we present generates the large itemsets by building the expression tree and prunes away the undesired one by checking the acceptance range. In our work, we consider constraints of arithmetic inequalities which are composed of common operators such as +, -, *, and /. Preliminary experimental results of the algorithm in comparison with the classical a-priori algorithm are also reported.