Split and Win Apportioning Algorithm – Swaa to Discover Frequent Patterns in Large Database

Abstract

Mining large datasets and discovering meaningful hidden patterns is not a new area but a lot of improvement is essential to overcome the cost and operational overheads, this paper finds a solution by splitting the large dataset finding the individual partition support count (IPSC) and then the partitioned dataset are merged to find the merged partitioned support count (MPSC) to reduce the burden of time and memory related issues. To find the IPSC and MPSC simple bit vector approach is utilized. The proposed algorithm is compared with the other existing algorithms to gauge its performance with respect to speed and the memory consumption.