An Emphasized Apriori Algorithm for Huge Sequence of Datasets

Abstract

In this paper, an Emphasized Apriori algorithm is proposed based on the most popular Apriori algorithm to overcome its drawbacks which are nothing but time consumption and the memory space. Since the Apriori algorithm scans the entire database based on minimum support and minimum confidence, it consumes more time and also more space. In this approach, new candidate set was prepared by considering the minimum support such that the total number of items will be reduced intern reduces the time taken and also memory space. This approach also includes the concept of Frequent Pattern (FP) growth to delete the items which are not frequent. Finally, a FP tree is established based on the frequent itemsets such that the items which are not frequent were delete intern to reduce the space consumption.