面向RDF数据的语义扩展查询方法  

Semantic Extension of Query for RDF Data

作　　者：李璞[1,2] 肖宝[3] 

机构地区：[1]郑州轻工业学院软件学院,郑州450000 [2]华南师范大学计算机学院,广州510631 [3]钦州学院电子与信息工程学院,广西钦州535000

出　　处：《河南大学学报（自然科学版）》2016年第4期453-461,共9页Journal of Henan University:Natural Science

基　　金：河南省基础与前沿技术研究计划项目(102300410110);广西高校中青年教师基础能力提升项目(KY2016LX431)

摘　　要：随着大数据时代的到来,互联网所承载的信息呈现出异构性强、价值密度低、来源分布广、实时性高等新的特点.因此在信息检索时,与获取相关网页相比,人们更希望获得页面中包含的知识.RDF作为一种新的知识表示和发布形式,以其精确且易于理解的语义结构得到了广泛的认可和应用.但传统基于语法匹配策略的SPARQL查询方法无法发现RDF数据中不同谓词间的隐含语义,不能从语义层面上获取更加完备的查询结果.针对这一问题,设计了一种新的查询方法来扩充传统SPARQL查询模式.首先分析RDF三元组中谓词的语义属性.在此基础上,根据不同谓词间语义属性的标注,对输入SPARQL语句中具有语义属性的谓词进行扩展,从而给出了一种新的一般化的查询策略.最后通过实验验证了这种语义扩展查询方法的可行性和正确性.With the advent of Big Data Era, the online information has features such as high heterogeneity, low density value, widely distributed sources, and real-time updating. So during information retrieval, users prefer to get knowledge rather than pages from Web. RDF, a new form of knowledge representation and publishing, can be considered as a general model for graph-structured data encoded as triples （also known as facts） in the form of subject-predicate-object （s, p, o）, and can provide a more precise and comprehensible semantic structure. The core idea of RDF Schema （RDFS） is to extend upon RDF vocabulary and allow attachment of semantics to user defined classes and properties. Further, SPARQL, as a standard query language of RDF, has been widely recognized and applied. The basic element of SPARQL query language is the triple pattern which contains one or more variables marked by the use of either ＂？＂ or ＂ $ ＂. A set of triple patterns consist of so-called basic graph patterns （BGP）. However, the bottleneck here is that the RDF terms in datasets may have some implicit semantics reflected by the features and properties of the relations which are called predicates in RDF triples, while the execution of SPARQL query aims at finding the matching triples from datasets whose forms are consistent with the query pattern in syntactic level. Obviously, traditional SPARQL-based querying approaches cannot fully express the semantics of the query for RDF data and achieve additional answers in semantic case. To fill this gap, in this paper, a new querying approach was designed to extend the SPARQL pattern. The semantic properties of predicates in RDF triples were analyzed. On this basis, a novel general strategy for executing queries by integrating semantic extension into the SPARQL protocol was given. This new query scheme semantically extends the predicates in input SPARQL statements by matching the annotations of the semantic attributes of predicates. Lastly, the feasibility and soundness of this se

关 键 词：语义扩展 RDF数据 大数据 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]