Discovery of Direct and Indirect Association Patterns in Large Transaction Databases

Abstract

Association rules mining is one of the important tasks in data mining research. While most of the existing discovery algorithms are dedicated to efficiently mining of frequent patterns, it has been noted recently that some of the infrequent patterns can provide useful insight view into the data. As a result, indirect association rules have been put forward, the traditional association rules are called direct association patterns. All the existing algorithms for mining indirect association rules need get all frequent itemsets using Apriori or other algorithms for mining association rules, then generate indirect association candidates using frequent itemsets. Instead of this method, we put forward an approach to discover both direct and indirect association patterns. Key words: Direct Association Pattern, Indirect Association Pattern, Data Mining