基于前缀项集的Apriori算法改进  

THE IMPROVEMENT OF APRIORI ALGORITHM BASED ON PREFIXED-ITEMSET

作　　者：于守健[1] 周羿阳 

机构地区：[1]东华大学计算机学院,上海201600

出　　处：《计算机应用与软件》2017年第2期290-294,共5页Computer Applications and Software

摘　　要：关联规则的挖掘是数据挖掘中一个重要内容,主要目的是找到事务数据库中的有趣的模式。Apriori算法是关联规则挖掘的最经典算法之一,但是它本身存在着效率上的瓶颈。在深入了解Apriori算法前提下,提出基于前缀项集的候选集存储结构,并利用哈希表在快速查找上的优势,大大提高了经典Apriori算法在连接步骤和剪枝步骤中的效率。实验证明改进后的Apriori算法在一定支持度下比经典Apriori算法有着更大的效率优势,并且支持度越小时提升效率越大。The mining of association rule is an important method for discovering interesting relations between variables in large databases. Apriori algorithm is one of the most classical algorithms of association rules,but it has bottleneck in efficiency. Thus,a candidate item set storage structure based on prefixed-item set is proposed with the help of the quick search of hash map,and the efficiency of classical Apriori algorithm in connecting and pruning step has been improved greatly. The experiments show that the improved Apriori algorithm does better in efficiency than the classical Apriori algorithm in certain degree＇s support,and the smaller support,the better efficiency.

关 键 词：数据挖掘 APRIORI算法 前缀项集 关联规则 哈希表 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]