Milling Up-to-Date Knowledge Based on Tree Structures

摘要

In the past, the up-to-date patterns is proposed to mine the frequent itemsets within its corresponding lifetime. This hybrid method is based on the Apriori-like approach, which requests high computational cost and memory requirement. In this paper, the up-to-date pattern tree (UDP tree) is proposed to keep the up-to-date patterns in a tree structure. The experimental results show that the proposed approach has a better performance than the level-wise up-to-date algorithm.