基于动态分组的增量式挖掘关联规则更新算法

数据挖掘的一个重要方面是挖掘关联规则，目前已提出了包括经典算法Apriori在内的许多算法，而在实际关联规则的挖掘过程中，用户将需要不断调整用于描述用户兴趣程度的阈值：最小支持度和最小置信度。如何维护已发现的关联规则变得至关重要。该文提出的GIUA算法解决了在数据库D不变的情况下，最小支持度和最小置信度发生变化时关联规则的维护问题，最大效率地利用原有结果，通过动态分组将连接步和修剪步的循环减到最少，并尽可能地将挖掘过程并行化。

Incremental Updating Algorithms for Mining Association Rules Based on Dynamic Grouping

Mining association rule is an important data mining problem.There have had many algorithems which include classical Apriori algorithms.However,in practical process of mining association rules,users have to continually tune two thresholds, minimun support and minimum confidence,which describe users special interesing, how to maintain discovered association rules has become very important.Assuming that database D is not updated,GIUA solves the problem of maintaining discovered association rules when minimum support and minimum confidence change,using the intrinsic result most effective, reduces the circle of the connection step and the shaving step least via dynamical grouping,and do it is possible to paralleling the mining process.