Mining Positive and Negative Weighted Association Rules from Frequent Itemsets Based on Interest

The weighted association rules (WARs) mining are made because importance of the items is different. Negative association rules (NARs) play important roles in decision-making. But the misleading rules occur and some rules are uninteresting when discovering positive and negative weighted association rules (PNWARs) simultaneously. So another parameter is added to eliminate the uninteresting rules. A new model in the paper of extending the support-confidence framework with a sliding interest measure could avoid generating misleading rules. An interest measure was defined and added to the mining algorithm for association rules in the model. The interest measure was set according to the demand of users. The experiment demonstrates that the algorithm discovers interesting weighted negative association rules from large database and deletes the contrary rules.

1. Introduction

Association Rule is an important type of knowledge representation revealing implicit relationships among the items present in large number of transactions. Association rule mining (ARM) is firstly proposed by R.Agrawal, T.Imielinski and A.Swam in 1993[1]. The fast algorithm called Algorithm Apriori is put forward in 1994[2]. Much effort has been devoted and algorithms proposed for discovering association rules efficiently [3] [4]. These conventional association rules are widely used in many application domains, such as market basket analysis and web trace.