基于数组向量的Apriori算法改进  

AN IMPROVED Apriori ALGORITHM BASED ON ARRAY VECTORS

作　　者：林佳雄[1] 黄战[1] 

机构地区：[1]暨南大学信息科学技术学院,广东广州510632

出　　处：《计算机应用与软件》2011年第5期268-271,共4页Computer Applications and Software

摘　　要：在分析Apriori算法的基础上,提出一种改进Apriori的算法,其主要思想是基于数组向量的数据布局,即根据每个事务不同的项目长度,分别存储在相应列长度的二维数组中,同时合并项目完全相同的事务,并记录事务的数目。新的算法改进了连接比较的次数、减少不必要事务的扫描和提高了算法对内存空间的利用效率。应用示例表明,该算法时间复杂度降低;实验结果显示算法是有效可行的。The paper proposes an improved Apriori algorithm based on the analysis of present Apriori algorithm,whose main thought is on the basis of data layout of the array vectors,i.e.different item lengths of different transactions to decide on which two-dimensional arrays of the corresponding column lengths to store in,at the mean time deleting repetitive transactions and counting the transactions.The improved algorithm decreases comparison times during joining operations,reduces unnecessary transaction scans and increases the memory use efficiency of the algorithm.The implementation of the improved algorithm demonstrates a decrease of time complexity.Experiment results prove the effectiveness and feasibility of the improved algorithm.

关 键 词：数据挖掘 关联规则 频繁项集 APRIORI算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]