基于项集特性的关联规则挖掘中Apriori算法的改进

提出一个改进Apriori的算法IApr,利用事务数据库的事务数相对于项集的项的个数而言要大得多这一特点,采用线性存储结构,并结合推出并证明的项集特性,考虑候选频繁项目集的各个项的应满足的条件,只需扫描一次事务数据库,有效减少了生成候选频繁项集的数目,从数据扫描量、搜索空间、时间复杂度上分析都提高了算法效率。

An Improved Apriori Algorithm for Mining Association Rules Based on the Properties of Item Sets

This paper presents an improved algorithm of IApr,because of the characteristics that transaction of transaction database are much bigger as to the item of item sets in number,the IApr uses linear data structure cooperating with the properties of item sets,and considers the condition that each item of candidate frequent item sets should be met,it scans the transaction database only one time so as to reduces the number of candidate frequent item sets,and improves the algorithm efficiency on the aspects of data scanning, searching space and time complexity.