一种支持用户偏好的RDF模糊查询方法

摘要

RDF fuzzy retrieval is an important module for realizing intelligent retrieval in Semantic Web.In this paper,Zadeh's type-II fuzzy set theory,as well as the concepts of α-cut set and linguistic variable was adopted to put forward the RDF fuzzy retrieval mechanism supporting user preference,which extends SPARQL to express fuzzy and preference conditions.Moreover,ordered sub-domain table of linguistic values was constructed to realize the projection from the fuzzy values to relayed sub-domains in the table,so as to figure out the interval of membership.On this basis,extended queries were then converted into standard SPARQL queries with a set of defuzzification rules,so as to achieve fuzzy retrieval operations.In order to test the ideology proposed in this paper,the fp-SPARQL retrieval system was developed.According to the result of this experiment,the method improves the performance of RDF fuzzy retrieval,and correspondingly,users' satisfaction rate on the retrieval results is also enhanced.%RDF模糊查询是实现语义Web智能检索的重要组成部分,利用Zadeh的Ⅱ型模糊集合理论、α-截集及语言变量概念,提出了支持用户偏好的RDF模糊查询方法,其扩展了SPARQL语言来实现模糊及偏好表达,构造了有序语言值子域袁来实现模糊值到相应子域的映射,以确定隶属度区间.利用去模糊化规则,将扩展的查询转换为标准SPARQL,利用现有的SPARQL查询引擎实现模糊查询操作.为验证提出的方法,开发了fp-SPARQL实验系统.实验结果表明,该方法提高了RDF模糊查询效率,增强了用户对查询结果的满意度.