摘要： 为了解决加权支持率可能大于1的不足,对属性的权重集归一化，并提出第1类加权关联规则挖掘算法。此算法能有效地考虑属性的权重，并且认为规则的重要性随着规则中所含属性数量的增加而增加；但在有些数据库中，挖掘关联规则只需考虑属性的权重，也就是说规则的重要性不随着规则中所含属性数量的增加而增加。该文通过对各个元素所作贡献的加权构造了一个加权数据库，提出了解决此类问题的第2类加权关联规则挖掘算法。

关键词: 数据挖掘, 权重, 支持率, 信任度, 关联规则

Abstract: For solving the problem that the weighted support may be bigger than 1, the weight set of attributes is normalized, and the first algorithm for mining weighted association rules is provided. This algorithm can effectively consider the importance of Boolean attributes, and consider that the importance of association rule is increased with the amount of attributes in the rule. But in some databases, the algorithm only needs to consider the importance of Boolean attributes, and does not need to consider that the importance of association rule is increased with the amount of attribute in the rule. To solve this problem, a weighted database is constructed using the weighted contribution of each element in this paper, and the second algorithm for mining weighted association rules is provided.

中图分类号: 

• TP311.12