相关系数和卡方检验的正负关联规则挖掘算法

随着经济全球化和信息技术的发展,为了给企业发展提供更多的信息支持和决策帮助,数据中心纷纷建立起来,其作用是通过分析海量数据来为企业的政策趋向和战略选择提供意见佐证.但是,要想在庞大的数据海洋中获取数据间的相关性依赖并非易事,而且,传统的关联规则算法通常并不完善,产生的规则通常会包含一些没有意义甚至错误的规则,即所谓的弱规则与负规则.针对这种现状,提出一种度量正负关联规则的检验方法,并引入赋予不同权重值给不同数据库的方式,提高在水平多数据库中挖掘正负关联规则的效率.

Mining of positive and negative association rules and development tendency in multi-database

As the developing of economy＇s globalization and information technology, many enterprises have across the country, and even in the world scope have their own data center.They use the association rules by analysis of these data for their strategic choices and policy making service.But, it＇s not an easy thing to search for relationship between these huge data by the directly way ,what＇s more,association rules generated by Apriori algorithm includes some useless and even misleading rules.This paper puts forward a kind of measuring method of positive and negative association rules,extend the algorithm by place different weight on dif- ferent database to make it more efficient to mine both positive and negative association in multi-database.