一种基于分布式rough本体的语义相似度计算方法  

A distributed computing method of semantic similarity based on rough ontology

作　　者：常宝娴[1] 陈玮玮[1] 李素娟[1] 

机构地区：[1]南京工业大学理学院,南京211816

出　　处：《扬州大学学报（自然科学版）》2014年第1期60-62,66,共4页Journal of Yangzhou University：Natural Science Edition

基　　金：江苏省高校自然科学基金资助项目(11KJB520006)

摘　　要：针对传统的语义相似度计算方法缺少相应领域本体和精确知识支撑等缺陷,提出一种基于分布式rough本体的语义相似度计算方法.通过半自动构建领域本体保证语义相似度计算的准确度,采用rough的上下近似提高语义相似度计算的精确性,并通过实例验证了该算法的有效性.结果表明:该文方法不仅可减弱对领域专家的依赖,而且还能大幅提高语义相似度计算的查全率和准确率.This paper presents a distributed computing method of semantic similarity based on rough ontology and improves the precision according to ontology, the completeness according to the upper approximation and tow approximation of rough theory. It also improves the independence according to distributed data processing. An experiment of gathering web pages automatically of remote edu- cation is used to construct domain rough ontology and compute the semantic similarity. The experiment shows that the algorithm not only reduces the dependence of domain experts, but also greatly enhances the rates of completeness and precision.

关 键 词：rough本体 语义相似度 分布式 远程教育 

分 类 号：TP311.51[自动化与计算机技术—计算机软件与理论]