Dynamic FP-Tree Pruning for Concurrent Frequent Itemsets Mining

To solve the problem of huge memory usage of FP-tree construction and traversal in FP-growth, Dynamic-prune, which is a concurrent frequent itemsets mining algorithm based on FP-tree is proposed. On the one hand, by recording the change of support counts of frequent items during the process of FP-tree construction, dynamic FP-tree pruning is implemented. And the rationality is discussed. On the other hand, by using the concurrency strategy, the construction of FP-tree and the discovery of frequent itemsets can be realized simultaneously. Compared with FP-growth, it is not necessary to mine frequent itemsets after the construction of the whole FP-tree in Dynamic-prune. Experimental results show Dynamic-prune is efficient and scalable.