上海大学学报(自然科学版)

• 通信与信息工程 • 上一篇    下一篇

基于本体核与直方图的聚类预处理方法

夏骄雄,徐俊,吴耿锋   

  1. 上海大学 计算机工程与科学学院,上海 200072
  • 收稿日期:2007-04-27 修回日期:1900-01-01 出版日期:2008-02-28 发布日期:2008-02-28
  • 通讯作者: 夏骄雄

Cluster Preprocessing on Ontic Kernel and Histogram

XIA Jiao-xiong,XU Jun,WU Geng-feng   

  1. School of Computer Engineering and Science, Shanghai University, Shanghai 200072, China
  • Received:2007-04-27 Revised:1900-01-01 Online:2008-02-28 Published:2008-02-28
  • Contact: XIA Jiao-xiong

摘要: 围绕数据对象与用户需求之间的关联特性,该文将“本体论”基本理念引入聚类分析过程,定义“本体核”和“客体数据”来表述数据对象在数据空间中的动态分布,并通过弱(强)量本体核构建数据对象间的相互关系.同时,利用“本体核”中客体数据的频数及其同频异值数构造聚类分析直方图,分别使用两种控制聚类合理性函数来确定直方图门限值的预处理算法,获取数据对象的最终聚类归属,并通过应用示例验证方法的可行性和有效性.

关键词: 本体核, 合理性函数, 聚类预处理, 直方图

Abstract: According to the connective characteristic of data and demands within data space, the principle of ontology is introduced into the process of cluster analysis. The pinpoint existence modality and dynamic distribution of data are described on the definition of ontic kernel and object data. Interrelation among the data can be studied with weak ontic kernel and strong ontic kernel. The cluster analysis histogram is constructed on the time of object data and different value on the same time. For different histogram statement, the algorithm that can process the limited gate value with rational function of controlling cluster is presented to determine the ultimate generic cluster. An application illustration is used to demonstrate feasibility and validity of the cluster preprocessing on ontic kernel and histogram (CPOKH). 

Key words: cluster preprocessing, histogram, rational function, ontic kernel

中图分类号: