模糊概念格在知识发现的应用及一种构造算法  

Research on Fuzzy Concept Lattice in Knowledge Discovery and a Construction Algorithm

作　　者：强宇[1] 刘宗田[1] 林炜[1] 时百胜[1] 李云[1] 

机构地区：[1]上海大学计算机学院

出　　处：《电子学报》2005年第2期350-353,共4页Acta Electronica Sinica

基　　金：国家自然科学基金<分布式概念格数学模型及算法研究>(No.60 2 750 2 2 )

摘　　要：基于有限L_背景的模糊格在扩展和时空复杂度上有局限 .本文定义了广义的模糊概念格和其上的截运算以简化格构造 ,提出了一种模糊格构造算法 .在概念格结点级上定义了两个模糊参数α和 δ ,以避免提取因高偏差导致的无效规则 .给出一个实例 ,说明了从模糊概念格提取不确定规则、计算规则支持度、置信度的原则、方法 .实现了构造算法与Godin算法的对比实验 ,结果表明本算法在时空性能上要优于Godin算法 .Fuzzy lattice based on finite L_context has limit on expansion and time and spatial complexity. Fuzzy-attributes concept lattice in common sense and cut calculation on it were defined to simplify lattice construction. A construction algorithm was presented. Two fuzzy parameters a and S on level of concept lattice node were defined to avoid extracting invalid association rules induced by high abbreviation. A case was given to demonstrate the principles for abstracting indefinite rules and computing their support and confidence. A comparison experiment between construction algorithm and Godin algorithm was made. The result shows that construction algorithm is superior to Godin algorithm on time-spatial complexity.

关 键 词：人工智能 属性模糊概念格 模糊语言变量 知识发现 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]