Academy & Industry Research Collaboration Center
Association rules are one of the data mining methods for discovering knowledge from large amounts of data in databases. In this paper, the authors propose an intelligent method for discovering association rules, called IMAR. IMAR is designed through three main phases, i.e., preprocessing, processing and post processing. It has been experimented using three domain data sets, i.e., Australian Credit Card (ACC), Jakarta Stock Exchange (JSX), and CLEVeland Heart Diseases (CLEV) data sets. The experimental results show that IMAR can discover association rules from large inconsistent databases intelligently and accurately, and reduce the number of generated interesting association rules without losing information and with higher accuracy.