The Mining Algorithm of Frequent Itemsets based on Mapreduce and FP-tree

The date mining based on big data was a very important field. In order to improve the mining efficiency, the mining algorithm of frequent itemsets based on mapreduce and FP-tree was proposed, namely, MAFIM algorithm. Firstly, the data were distributed by mapreduce. Secondly, local frequent itemsets were computed by FP-tree. Thirdly, the mining results were combined by the center node. Finally, global frequent itemsets were got by mapreduce and the search strategy. Theoretical analysis and experimental results suggest that MAFIM algorithm is fast and effective.