基于有向边和属性的相似度模型设计  

Similarity Model Design Based on Edge and Attribute

作　　者：崔岩[1] 张淑丽[1] 

机构地区：[1]西北工业大学明德学院计算机信息技术系,西安710124

出　　处：《微处理机》2014年第5期47-50,55,共5页Microprocessors

摘　　要：通过分析单一条件下相似度算法的计算原理及设计思想,以有向图的形式描述概念,分析和推导了有向边、语义重合度及属性相似度的计算模型,给出了一个含有语义及属性权值的概念相似度计算模型。最后通过对比传统模型的性能测试,验证了新模型对相似度计算的精度。Through the analysis of the calculation principle and design concept of similarity algorithm under the single condition,the concept of directed graph is described.The calculation models of the directed edge,the semantic overlap and the attribute similarity are analyzed and concluded to establish the concept similarity calculation model with semantic and attribute weight.Finally,the accuracy of the similarity calculation of the new model is verified by comparing with the traditional one.

关 键 词：概念相似度 有向边 属性权值 本体 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]