摘要
当形式背景中的某些属性可以分解为若干子属性时,构建分层概念格模型,讨论原形式背景的概念格和分层形式背景的分层概念格之间的联系.提出由下层概念构造上层概念的上卷构造算法和由上层概念构造下层概念的下钻构造算法.实验表明,采用分层概念格模型可以建模复杂属性数据,分层概念格上卷构造算法和下钻构造算法可以提高构建概念格的效率.
A model of layered concept lattice is established,when the attributes of a formal context can be decomposed into some sub-attributes. The relationship between the original concept lattice and the layered concept lattice is discussed. Two algorithms are proposed: the roll-up algorithm and the drill-down algorithm. In the roll-up algorithm,the upper concept is constructed by the lower concept,and in the drill-down algorithm,the lower concept is constructed by the upper concept. Examples and numerical experiments show that the layered concept lattice model can be used to model complex attribute data.Furthermore,the roll-up algorithm and the drill-down algorithm improve the efficiency of building concept lattice.
作者
张家录
吴霞
钟嘉鸣
陆汝华
ZHANG Jialu1, WU Xia1 ,ZHONG Jiaming2 , LU Ruhua3(1. College of Mathematics and Finance, Xiangnan University, Chenzhou 423000 ;2. College of Economic and Management, Xiangnan University, Chenzhou 423000; 3. College of Software and Communication Engineering, Xiang- nan University, Chenzhou 42300)
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2018年第4期310-321,共12页
Pattern Recognition and Artificial Intelligence
基金
湖南省自然科学基金项目(No.2017JJ2241
2016JJ6138)
湖南省社会科学基金项目(No.16YBA329
13YBB205)~~
关键词
形式背景
粒计算
分层概念格
上卷构造算法
下钻构造算法
Formal Context
Granular Computing
Layered Concept Lattice
Roll-Up Building Algorithm
Drill-Down Building Algorithm