摘要
装备元数据进行检索、交换时需要计算相似度,主要对相似度算法进行设计。装备元数据分解为概念和属性两种元素进行元素相似度计算,概念相似度由名称、结构和实例三部分的相似度计算叠加而成,属性相似度则综合考虑名称、定义域和值域的相似度。对各个部分的相似度计算设计了具体的计算公式。装备元数据之间的相似度在元素相似度基础上叠加计算而得出。这种算法能够真实地反映元素和元数据间的相似程度,且实用可行。
When the Equipment Metadata is searched and exchanged, the similarity calculating is required. This paper mainly designs the similarity algorithm. The Equipment Metadata is decomposed into two elements of concept and attribute for similarity calculating. Concept similarity composed of name, structure and examples similarity calcu- lating. Attribute similarity takes the similarity calculating of name, domain of definition and domain of value into ac- count. Specific formulas are designed for various similarity calculating parts in the paper. The Equipment Metadata mutual similarity is integrated from element similarity. This algorithm can truly reflect the similar degree of metadata and elements, and is feasible and useful.
出处
《计算机仿真》
CSCD
北大核心
2010年第4期84-86,95,共4页
Computer Simulation
基金
总装备部重点资助项目(编号:2006RW37)
关键词
装备元数据
相似度
概念相似度
属性相似度
Equipment metadata
Similarity
Concept similarity
Attribute similarity
