一种高效的关联规则更新算法

在大型数据库中 ,数据不是静止的 ,新的记录需要不断地增加到旧的数据库中 ,而从旧的数据库中挖掘出来的关联规则必须加以维护 ,为此提出了一种关联规则更新算法。该算法既能有效地利用已经发现的知识 ,又能减少数据库的扫描次数 ,同时只需产生少量候选集。实验表明 ,该算法的执行效率比现有的典型更新算法高 ,是一种有实际应用价值的、高铲的关联规则更新算法。

An Efficient Updating Algorithm for Association Rules

Data are not static in large databases.When new data are added to the old database,association rules mined have to be maintained. In this paper,an updating algorithm for association rules is proposed,which does not only utilise the knowledge available,but also reduce the times of database scanning,and only a small number of candidate sets are needer. The experimental result shows that the running efficiency of the updating algorithm is higher than the existing algorithms. Therefore,the algorithm is practical and efficient.