Improved Frequent Pattern Mining Algorithm using Divide and Conquer Technique with Current Problem Solutions

Frequent patterns are patterns such as item sets, subsequences or substructures that appear in a data set frequently. A Divide and Conquer method is used for finding frequent item set mining. Its core advantages are extremely simple data structure and processing scheme. Divide the original dataset in the projected database and find out the frequent pattern from the dataset. Split and Merge uses a purely horizontal transaction representation. It gives very good result for dense dataset. The researchers introduce a split and merge algorithm for frequent item set mining. There are some problems with this algorithm. We have to modify this algorithm for getting better results and then we will compare it with old one. We have suggested different methods to solve problem with current algorithm. We proposed two methods (1) Method I and (2) Method II for getting solution of problem. We have compared our algorithm with the currently worked algorithm SaM. We examine the performance of SaM and Modified SaM using real datasets. We have taken results for both dense and sparse datasets.