Frequent itemsets mining algorithm based on node linked list FP-Tree

Aiming at the problem of low efficiencies for frequent item sets mining based on FP-Tree, which need a great lot of recursion call. This paper proposes another mining algorithm which uses a improved data structure named node linked list FP Tree (NLFP-Tree). NLFP-Tree compresses the FP-Tree greatly by record the node´s prefix path using node linked list. And mines frequent item sets by bottom-to-up and non-recursion method. Experimental results show that the mining algorithm performs better than Apriori, FP-Growth.