An algorithm based on predicate path graph for mining multidimensional association rules

Abstract

Multidimensional association rules is more common in association rule mining. For a shortcoming that current multidimensional association mining algorithms need to scan database many times, this paper presents an association rule mining algorithm, Ex-Apriori, based on the predicate path graph. The algorithm can produce the predicate path graph by scanning database only once, and dig out the frequent pattern based on the frequent predicate path graph. Hence it avoids the shortcoming of scanning database many times. The experiment shows that compared with the classical association rule mining algorithm Apriori, Ex-Apriori algorithm has a significant improvement in time efficiency. © 2013 Springer-Verlag.