Sandwich-Apriori: A combine approach of Apriori and Reverse-Apriori

The problem of generating large frequent itemset for the generation of association rules in the transactional database is considered. Previous work in this field already proposes many algorithms like Apriori, FP-growth, and their variations. Reverse-Apriori which is also a variation of Apriori for finding large frequent itemset in reverse manner, it has its own advantages and limitations like Apriori. This paper proposes a new approach called Sandwich-Apriori which is a combination of both Apriori and Reverse-Apriori. This Approach reduces number of scans and number of candidates generated as compared to Apriori and Reverse-Apriori. The worst case number of scans of the proposed algorithm is half of the Apriori. This approach gets pros of both algorithms with a new pruning technique to reduce the number of scans and candidates generated.