基于不可约元的概念格属性特征识别方法  

A Recognition Method of Attributes in Concept Lattices Based on Irreducible Elements

作　　者：李鸿儒[1,2] 魏平[3] 

机构地区：[1]西安交通大学理学院信息与系统科学研究所 [2]烟台大学数学与信息科学系烟台264005 [3]烟台大学数学与信息科学系

出　　处：《计算机科学》2006年第6期175-178,共4页Computer Science

基　　金：国家973计划资助(项目号:2002CB312200)。

摘　　要：形式概念分析是数据挖掘与知识获取的一种有效的数学工具。概念格作为形式概念分析的一个核心内容,刻画了对象和属性之间的内在联系。基于概念格中外延、内涵闭系统中不可约元的性质,本文研究了概念格理论中的属性分类问题,给出了一种属性特征的识别方法。这种方法不仅揭示了交不可约元与属性特征的关系,同时为知识约简提供了一种新的途径。Formal concept analysis is an effective mathematic tool for data mining and knowledge acquisition. Concept lattice is a key issue of formal concept analysis, and characterizes the essential relation of objects and attributes. Based on the properties of irreducible elements in the extent and intent closure systems, this paper investigates the problem of attribute classification in concept lattices and presents a recognition method of the characteristic of attributes. The method proposed not only reveals the relationships between the irreducible elements and the characteristics of attributes, but also provides a new way for knowledge reduction.

关 键 词：形式背景 概念格 闭系统 不可约元 知识约简 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]