具有数量特征的利润约束关联规则发现  

Discovery of Quantitative Profit Constraint Association Rules

作　　者：王立敏[1] 高学东[1] 武森[1] 

机构地区：[1]北京科技大学经济管理学院,北京100083

出　　处：《微计算机应用》2008年第3期1-5,共5页Microcomputer Applications

摘　　要：关联规则发现是数据挖掘的重要研究内容之一,具有广泛的应用。在以前的研究中,关联规则发现算法一般都没有考虑项目的销售数量。而在实际中,相关联项目的销售总利润对于商家来说非常重要。因此,在关联规则发现中引入销售数量的利润约束问题显得很必要。文章针对此问题提出了项集事务利润和事务利润的概念,并对项集利润和支持期望的概念进行了重新定义,接着给出了具有数量特征的利润约束关联规则挖掘问题的性质和定理,最后根据该问题的性质给出了高效的求解算法。Discovery of association rules is an important research topic in data mining, which has been found very useful in many applications. In previous work, items sales had not been considered in the algorithm discovery association rules. However, in fact, the profit on sales of correlational items is highly important for businessman. Therefore, it is necessary apparently that profit constraint in association rules mining has been induced sales. The paper discussed the question, present transaction profit of items concept and transaction profit concept. , Items profit and support expectation have been redefined, then presents the characters of association rules mining with quantitative profit constraint, finally proposes an effective algorithm according to the character of the problem.

关 键 词：数据挖掘 关联规则 数量特征利润约束 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] S812[自动化与计算机技术—计算机科学与技术]