Research on mining global maximal frequent itemsets for health big data

摘要

Traditional mining algorithms did not suit mining of global maximal frequent itemsets. Therefore, a new mining algorithm of global maximal frequent itemsets for health big data, namely, NMAGMFI algorithm was proposed. Firstly, the global frequent items were mined. Secondly, local FP-tree was reconstructed by each node. Thirdly, the mining results were combined by the center node. Finally, the global maximal frequent itemsets are mining by the strategy of top-down and FP-tree. Experimental results suggest that NMAGMFI algorithm is fast.