Research on Mining Maximum Frequent Itemsets Based on JFP-Growth Algorithm
Download as PDF
DOI: 10.23977/meet.2019.93706
Author(s)
Wang Zeru, Wang Hongmei
Corresponding Author
Wang Zeru
ABSTRACT
The FP-Growth algorithm is the most representative frequent item set mining algorithm. An improved algorithm JFP-Growth algorithm is proposed for the shortcomings of FP-Growth algorithm. When mining the maximum frequent itemsets, the JFP-Growth algorithm traverses the support of the first 1-item set and the 2-item set of the first-time data set statistics, and uses the frequent 2-item set as the pruning condition for full pruning and The merged nodes make there is no non-potential candidate 3-item set in the JFP-tree, and the conditional pattern base is generated without traversing the project header table during the mining process. Finally, the FP-Growth algorithm and DMFIA algorithm are compared in the mining results, the number of nodes in FP-tree and JFP-tree, mining efficiency, etc., which verifies the correctness and efficiency of the JFP-Growth algorithm proposed in this paper.
KEYWORDS
Frequent Itemsets, Fp-Growth Algorithm, Frequent 2-Item Set, Pruning, Two-Dimensional Count Table