PARALLEL MINING OF FREQUENT MAXIMAL ITEMSETS USING ORDER PRESERVING GENERATORS

In this paper, we propose a parallel algorithm for mining maximal itemsets. We propose POP-MAX (Parallel Order Preserving MAXimal itemset algorithm), a fast and memory efficient parallel algorithm which enumerates all the maximal patterns concurrently and independently across several nodes. Also, POP-MAX uses an efficient maximality checking technique which determines the maximality of an itemset using less number of items. To enhance the load sharing among different nodes, we have used round robin strategy which achieves load balancing as high as 90%. We have also incorporated bit-vectors and numerous optimizations to reduce the memory consumption and overall running time of the algorithm. Our comprehensive experimental analyses involving both real and synthetic datasets show that our algorithm takes less memory and less running time than other maximal itemset mining algorithms.