An Improved Association Rules Mining Algorithm Based on Power Set and Hadoop

Abstract

The association rules mining has an very important impact in data mining. As the rapid growth of datasets, the required memory increase seriously and the operating efficiency declines rapidly. Cloud computing provides efficient and cheap solutions to analyze and implement the association rules mining algorithms in parallel. This paper proposes an improved association mining algorithm based on power set and MapReduce programming model, which can process massive datasets with a cluster of machines on Hadoop platform. The results of the numerical experiments show that the proposed algorithm can achieve higher efficiency in the association rules mining.