基于散列技术的高效剪枝关联规则挖掘算法  

High Efficiency Pruning Algorithm For Association Rule Mining Based On Hash Technique

作　　者：彭永供[1] 王靓明[1] 朱敏[1] 段隆振[1] 

机构地区：[1]南昌大学信息工程学院,江西南昌330031

出　　处：《南昌大学学报（理科版）》2009年第5期494-498,共5页Journal of Nanchang University(Natural Science)

基　　金：江西省自然科学基金资助项目(2008GZS0074)

摘　　要：该算法结合项集的有序特性有效压缩了频繁项集的数量,提高了算法效率,但是频繁2-项集的产生仍需要占用大量的存储空间,本文利用散列函数对候选频繁2-项集进行剪枝,这样频繁集的数目大大减少,不仅节省了存储空间、减少了计算开销,而且又保持了频繁项集的完整性。It combines the ordinal character of itemsets to reduce the number of the frequent itemsets in order to improve the efficiency of algorithm. But the frequent 2 - itemsets still require much store space, it proposes a new method of pruning to improve the algorithm, that is, hash function id applied to prune candidate frequent 2 - itemsets, and the number of the frequent itemsets can be greatly reduced . So, not only the cost of computing is reduced and the store space is saved, but also the completness of frequent itemsets are kept.

关 键 词：关联规则 APRIORI性质 散列函数 项集有序 频繁项集 

分 类 号：TP311.12[自动化与计算机技术—计算机软件与理论]