管理评论

›› 2012, Vol. 24 ›› Issue (3): 108-114.

• 市场营销 • 上一篇    下一篇

基于意外度的关联规则深层知识发现及应用研究

李 军1,2,4,黄安强3,张玲玲1,石 勇1,2   

  1. （1.中国科学院研究生院管理学院，北京100190；2. 中国科学院虚拟经济与数据科学研究中心，北京100190；3. 北京航空航天大学经济管理学院，北京 100190；4. 英大泰和财产保险股份有限公司，北京 100005）
  • 收稿日期:2012-09-26 修回日期:2012-09-26 出版日期:2012-03-25 发布日期:2012-09-27

Research on the Formation Mechanism of Service Recovery Paradox

Li Jun1，2，4 , Huang Anqiang3, Zhang Lingling1 and Shi Yong 1,2   

  1. (1.Management School of Graduate University of Chinese Academy of Sciences, Beijing 100190;2.Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing 100190;3.School of Economy and Management of Behang University, Beijing 100091; 4.Yingdataihe Property Insurance Co., LTD., 100005 Beijing)
  • Received:2012-09-26 Revised:2012-09-26 Online:2012-03-25 Published:2012-09-27

摘要： 为了弥补传统关联规则挖掘产生大量冗余规则、难以直接用于决策支持的不足，本文提出了一种基于用户已有知识的规则意外度评价方法，并在此基础上设计了基于意外度的深层关联规则挖掘算法。算法的优点在于能够将用户已知的规则作为领域知识加入到数据挖掘过程从而有效过滤和已知规则相近的冗余规则，并且可以将新得到的规则加入知识库中实现知识的积累和重用。最后本文采用一个商场数据验证了该算法的有效性，并且对具有回馈模式的关联规则在商品促销中的作用进行了分析。

关键词: 意外度, 关联规则, 商品促销

Abstract: This paper designs a method of evaluating association rule expectedness and applies it to the new association rule mining algorithm which is called Association Rule Mining Based on Unexpectedness. This new algorithm has two advantages: first, this algorithm can effectively screen those redundant and useless rules; second, this algorithm is able to add new association rules into knowledge base to accumulate and reuse knowledge. This paper validates the efficiency of the algorithm by a test of a data set of commodity, and then analyzes the use of association rules with a feedback pattern.

Key words: association rule, expectedness, product promotion