负增量式关联规则更新算法

Negative Incremental Updating Algorithm for Maintaining Association Rules

模式维护是数据挖掘中一个具有挑战性的任务。现有的增量式关联规则更新算法主要解决两种情况下的维护问题:一是最小支持度不变,而数据量增加;二是数据量不变,而改变最小支持度。本文提出了一种负增量关联规则更新算法。实验表明,该算法是有效的。

Pattern maintenance is a challenging task in data mining. Existing incremental updating algorithms are designed for maintaining association rules in two ways. One is to maintain association rules with a constant minimal support while new data is added to original database. Another is to maintain association rules with different minimal supports and a constant database size. In this paper, a new algorithm, called Negative Incremental Updating Algorithm （NIUA）, is proposed for maintaining association rules. We experimentally evaluate the proposed approach, and illustrate that our algorithm is efficient and oromising.