基于矩阵的增量式关联规则挖掘算法

Incremental updating algorithm based on matrix for mining association rules

关联规则是数据挖掘的重要研究内容之一。针对数据库数据增加的同时最小支持度发生改变的关联规则更新维护问题,提出了一种基于矩阵的增量式关联规则挖掘算法IUBM。该算法采用简单的数组和位运算,在执行关联规则的更新时,既不用多次扫描数据库,也不产生庞大的候选项集。实例表明,该算法的时间复杂度和空间复杂度大大降低。

Association rules is an important part of data mining research.This paper proposes the improved incremental updating algorithm based on matrix named IUBM （Incremental Updating Algorithm based on Matrix）,in order to deal with the updating problems of association rules when the transaction database increases and the minimum support changes.Arrays and bit operation are adopted by the algorithm,the database will not be scanned for many times and the enormous candidate itemsets will not be produced.It＇s proved that the time complexity and space complexity of the algorithm are effectively reduced.