一种新的关联规则发现算法及应用研究

A New Effective Algorithm for Mining Association Rules in Large Database

文章对关联规则发现中应用较多的Apriori算法和DHP算法进行了分析,提出了一种新的关联规则发现算法。为了计算项目集的支持度,提出了数据库特征矩阵和特征矢量的概念。该算法通过对特征矢量求内积得出项目集(item-set)的支持度,只需遍历数据库一次,大大减少了对数据库的操作,提高了算法的效率。算法分析和实例证明验证了算法的快速性和高效性。

On the base of analysis of Apriori algorithm and DHP algrithm which is widely applied.To calculate the support of an item-set the authors define characteristic matrix and characteristic vector of debase.In this paper the au-thors present a new algorithm,which traverses the database just once,and eliminates the unnecessary trouble of travers-ing database.Experiments and algorithm analysis identifies its efficiency.