Item Based Remilitarization of Positive and Negative Association Rules

Abstract

One of the important research topic in data mining is association rule mining and it is focusing on developing association rule mining algorithms to find positive association rules effectively. Recently the research in association rule mining is concentrated on finding negative association rules, which can provide valuable information to the user. In this paper, a new approach is proposed to generate efficiently both positive and negative association rules from the transactional databases. A novel structure Item based Bit Pattern is used to utilizing less memory to reduce of database scans. In the process of generation of a rule a statistical measure correlation coefficient is considered as rule interestingness measure.Huge  number of rules can be discovered. Thus it becomes difficult for decision makers to find out the relevant  rules . item based refilterization is used  for relevant rules.  The method has been evaluated using synthetic databases and the experimental results show the efficiency and effectiveness. Keywords: Positive Association Rule, Rule Interestingness, Negative Association Rule