基于Apriori算法的加权关联规则的挖掘  

Mining of weighted association rules based on algorithm Apriori

作　　者：张秋余[1] 曹华[1] 

机构地区：[1]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《兰州理工大学学报》2007年第6期69-71,共3页Journal of Lanzhou University of Technology

基　　金：国家科技支撑计划资助项目(2006BAF01A21)

摘　　要：关联规则挖掘主要用来发现数据库中存在的频繁项集.利用权值标识项目的重要程度,提出一种新的关联规则——加权关联规则的挖掘.由于项目权值的引入,Apriori性质不再成立,频繁项集的子集不再一定是频繁的.为此,提出k-最小支持数的概念,对原有Apriori算法进行改进.该算法能够挖掘出现频率小但是带来更大利润的项目,使得挖掘出的关联规则更加满足决策者的需求,也更加符合实际需要.Association rules mining is mainly used to find frequent item sets in database. By taking weight value as a mark of the importance of individual item, mining with a new association rule-weighted association rule was proposed. Due to the introduction of this items weight, the truth of Apriori would not hold further. The subset of frequent item set would not also be exactly frequent. Thus, a concept of k-support minimum value of item sets was set forth, and an algorithm to discover weighted association rules was proposed. Using this approach, the items with low frequency and high profit could be mined, and the association rules were mined more to meet the needs of decision makers, and also more meet the practical needs.

关 键 词：数据挖掘 加权关联规则 加权支持度 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]