Mining Weighted Frequent Patterns Based on an Improved Weighted FP-tree

摘要

FP-tree is used to store crucial, quantitative information about frequent patterns in FP-growth algorithms, a classic algorithm of mining frequent patterns. But it does not meet for mining weighted frequent patterns. In this paper the limitations of the existing algorithms for mining weighted frequent patterns in some other papers are analyzed, the structure of a traditional FP-tree is improved and a new Weighted FP-tree (WFP-tree) is constructed. Then an effective algorithm to handle the problem of mining weighted frequent patterns is proposed and analyzed through an example. Experiment results show that the new algorithm is effective for large database.