摘要
形式概念分析中多粒度形式背景存在两种表示方法:一种是由多粒度标记属性诱导出的多粒度形式背景,另一种是由属性粒化诱导出的多粒度形式背景.为了对这两种表示方法进行对比研究,本文基于布尔属性取值分布提出了类属性块的概念,在此基础上定义可类属性块分割的形式背景,表明了当原始形式背景为可类属性块分割时多粒度形式背景的两种表示方法能够相互转化.可类属性块分割的研究方法为进一步讨论多粒度形式背景奠定了基础.
In formal concept analysis, there are two representation methods for multi-granularity formal contexts: one is based on multi-granularity labeled attribute, and the other on attribute granulation.In order to make a comparison of these two representation methods, this paper put forward the notion of class-attribute block from the viewpoint of distribution of values of Boolean attributes, and then the divisible formal context was defined by class-attribute block.Furthermore, it was demonstrated that the two representation methods of multi-granularity formal contexts can be converted into each other when the original formal context is divisible based on class-attribute block.The research method of class-attribute block division lays the foundation for the further study of multi-granularity formal contexts.
作者
刘之茗
李金海
LIU Zhi-ming;LI Jin-hai(Data Science Research Center,Kunming Universityof Science and Technology,Kunming 650500,China;Faculty of Science,KunmingUniversityof Science and Technology,Kunming 650500,China)
出处
《海南热带海洋学院学报》
2019年第5期59-64,共6页
Journal of Hainan Tropical Ocean University
基金
国家自然科学基金项目(61562050,61573173)
关键词
形式概念分析
概念格
多粒度标记属性
属性粒化
多粒度形式背景
formal concept analysis
concept lattice
multi-granularity labeled attribute
attribute granulation
multi-granularity formal context
