Computer Engineering and Applications ›› 2008, Vol. 44 ›› Issue (34): 149-151.DOI: 10.3778/j.issn.1002-8331.2008.34.046

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

Similarity measurement based on decomposition of complex object

CHEN Zhi-ping1,2   

  1. 1.Department of Computer,Fujian University of Technology,Fuzhou 350014,China
    2.Department of Computer Science and Technology,Tsinghua University,Beijing 100084,China
  • Received:2007-12-19 Revised:2008-03-24 Online:2008-12-01 Published:2008-12-01
  • Contact: CHEN Zhi-ping

基于复杂对象分解的相似性计算方法

陈治平1,2   

  1. 1.福建工程学院 计算机系,福州 350014
    2.清华大学 计算机科学与技术系,北京 100084
  • 通讯作者: 陈治平

Abstract: The similarity of complex objects can’t be easily computed using common similarity measurements.A new method based on decomposition of complex object is presented in this paper.Using relationship among attributes,the complex object structure is iteratively decomposed till the similarity of objects using the decomposed structure can be computed using the common measurements.Thus a tree-structured object structure can be achieved in decomposition.Using this tree structure,similarity is computed at different levels from bottom to up,and the similarity of root node,complex object,can be finally acquired.Its successful application in a big telecom enterprise shows that the new method has good performance.

Key words: date mining, clustering algorithm, similarity computing

摘要: 针对目前常用的相似性度量方法难以满足复杂对象的相似性计算要求,提出了一种基于复杂对象结构分解的分层相似性度量方法。根据属性间的关系紧密程度将复杂对象结构迭代分解,直至基于分解后的简单对象结构的对象可以使用传统相似性度量方法计算对象间的相似性。分解过程可以得到树型结构的对象结构划分。在此基础上,利用树型的对象结构从叶节点向根方向对对象逐层进行相似性度量,最终得到复杂对象的相似性综合度量。结合某大型电信运营商的套餐数据进行应用分析,证明该方法具有较好的性能。

关键词: 数据挖掘, 聚类算法, 相似性计算