关键词: 数据挖掘, 最大频繁项目集, 关联规则, 频繁模式树

Abstract: Mining association rule is an important matter in data mining, in which mining maximum frequent itemsets is a key problem in mining association rule. Many of the previous algorithms mine maximum frequent itemsets by producing candidate itemsets firstly, then pruning. But the cost of producing candidate itemsets is very high, especially when there exist long patterns. In this paper, the structure of a FP-tree was improved, a fast algorithm DMFIA-1 based on FP-tree for mining maximum frequent itemsets was proposed, which did not produce maximum frequent candidate itemsets and was more effective than DMFIA. The new FP-tree is a one-way tree and there is no pointer pointing its children in each node, so at least one third of memory is saved.

Key words: data mining, maximum frequent itemset, association rule, FP-tree