Visual Mining of Market Basket Association Rules

Abstract

AbstractData mining is increasingly becoming important in extracting interesting information from large databases. Many industries are using data mining tools for analyzing their vast databases and making business decisions. Mining association rules is an important data mining method where interesting associations or correlations are inferred from large databases. Though there are many algorithms for mining association rules, these algorithms have some shortcomings. Most of these algorithms usually find a large number of association rules and many of these rules are not interesting in practice. Hence, there is a need for human intervention in mining interesting association rules. Moreover, such intervention is most effective if the human analyst has a robust visualization tool for mining and visualizing association rules. In this paper we present a three-step visualization method for mining market basket association rules. These steps include discovering frequent itemsets, mining association rules and finally visualizing the mined association rules. Most previous visualization methods have concentrated only on visualizing association rules that have been already mined by using existing algorithms. Our method allows an analyst complete control in mining meaningful association rules through visualization of the mining process.KeywordsInformation VisualizationAssociation ruleMarket BasketData Mining