Study of Distributed Mining Association Rules Based on Meta-learn Technology
Shuping Yao · Acta Simulata Systematica Sinica · 2004
Mining association rules is an important task of data mining. In this paper, a method and algorithm DMAR of distributed mining association rules is presented in distributed transaction database by using meta-learn technology. The algorithm has higher efficiency of mining and lower amount of communication. Distributed factor g is defined for measuring efficiency of mining algorithm, and it is pointed out that mining efficiency and data communication amount is related to g value. In the end, algorithm DMAR is proven to be correct and effective by experiment.