Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (7): 171-174.DOI: 10.3778/j.issn.1002-8331.2009.07.051

• 数据库、信号与信息处理 • Previous Articles     Next Articles

Mining algorithm base on multilevel weighted update association rule

SHI Fu-gao,YANG Feng-jie   

  1. College of Computer Science and Technology,Jilin University,Changchun 130012,China
  • Received:2008-01-21 Revised:2008-04-17 Online:2009-03-01 Published:2009-03-01
  • Contact: SHI Fu-gao

基于多级加权更新关联规则的挖掘算法

师富杲,杨凤杰   

  1. 吉林大学 计算机科学与技术学院,长春 130012
  • 通讯作者: 师富杲

Abstract: The date-set are managed equally and conformably in the traditional association rule mining algorithm.This paper brings forward a kind of increment association rule mining algorithm:introduce a update association mode with multilevel weighted,the date are endowed with weights during mining algorithm,enhances the effect for the mining result.The refresh date are embodied with higher interest gene for currently decision making.The test indicates the algorithm with high sensitivity responding for the direction and providing a timely and exactly information for the decision-making person.

摘要: 传统的关联规则挖掘算法对更新的数据集按平等一致的方式加以处理。提出了一种新的增量关联规则挖掘算法:引入了多级加权的更新关联规则模式,对最近更新的数据集在挖掘算法中赋予较高的权值,提高其对挖掘结果的影响力度。使得最近更新数据对当前决策拥有更高的兴趣因子得以充分体现。实验表明,该算法拥有较高灵敏度,并能及时响应事务集的发展趋势,从而为决策者提供及时、准确的信息。