Abstract: Aiming at the shortage of Apriori algorithm in find of frequent items such as numerous search-designate database set too many times and gennerate too many candidate itemsets, this paper proposes the TT-Apriori algorithm. This algorithm maps the tings-datebase into transaction-tree. Using the transaction tree can quickly find the maximal frequent itemsets. In the meamwhile it can simplify the calculation of support and avoid the scanning of the entire database and a large number of connecting steps to improve the efficiency of the mining.

Key words: maximal frequent itemsets, TT-Apriori algorithm, transaction tree, vector in plot

摘要： 针对Apriori算法在寻找频繁项集的过程中需多次扫描数据库、侯选项集过多、支持度计算过于复杂等问题，提出TT-Apriori算法。该算法将事务数据库转化成事务树，通过遍历事务树能直接快速地找到最大频繁项目集。简化支持度的计算，避免对整个数据库的扫描和大量的连接步骤，从而提高挖掘效率。

关键词: 最大频繁项集, TT-Apriori算法, 事务树, 向量内积

CLC Number: 

• TP311.52