基于关系数据库的RDF数据存储
学 校： 院 系： 专 业： 班 级： 姓 名： 学 号： 指导教师： 教师职称：
完成日期：
摘要
语义网的发展，引发了对RDF(Resource Description Framework,资源描述框架)的研究热潮。RDF 用来描述语义网中的资源，它是用来描述元数据的数据。随着RDF应用的发展，对于海量RDF数据的存储和查询也提出了更高的要求。W3C组织提出了的SPARQL查询语言草案，被视为未来语义网查询语言的发展方向。SPARQL 的半结构化查询方式与 RDF 的半结构化特性完美地结合起来。 本论文对RDF与SPARQL进行了深入研究，设计了使用关系数据库来进行RDF数据存储与查询的解决方案方案。
存储方面，将RDF数据结合SPARQL查询语言的格式设计存储结构，方便查询语言进行匹配，利用编码机制存储数据节省了存储空间。
查询方面，将用户提交的SPARQL查询语言转换成等价的SQL语言，提交给关系数据库进行查询，最后将查询结果返回给用户。本文详细分析了SPARQL 基本图模式，可选图模式，值限制图模式和并图模式，并针对现有的关于查询图模式匹配顺序的分歧，提出了自己的观点以及论据。最后采用查询操作树将 SPARQL查询语言正确转换成SQL语句，并提出了值限制图模式的转换算法。
关键词：RDF、SPARQL、SQL、存储结构、数据处理
ABSTRACT
With the development of semantic Web,worldwide attention has been profoundly obtained to the research boom of RDF(Resource Description Framework)that is the data for describing metadata and is used to describe Web resources in semantic Web. The storage and query of massive amount of RDF data,as the RDF application developed,demands for higher requirements.The draft of SPARQL query language provided by W3C organization is considered as the future development trend for semantic Web query language.The semi-structured query methods is able to be perfectly combined with the semi-structured characteristics of RDF, of SPARQL. This paper is deeply engaged in the research of RDF and SPARQL has designed solutions of utilizing Relation Database for storing and querying RDF data. In storage, the pattern of designing storage structure of combining RDF data with SPARQL query language is able to facilitate the matching for query language. Encoding mechanism used to store data saves storage space.
In query, the SPARQL query language submitted by users is initially transformed to equivalent SQL language and subsequently delivered to Relation Database for query, finally the query results have been returned to the users. This paper analyzes the basic SPARQL graph patterns, optional graph patterns, map mode and restricted mode.Own opinion as well as its argument has been proposed by detailed analyzing the structure of query map mode and aiming at the recent differences corresponding to the matching sequences of query map mode.Ultimately, Operation tree has been introduced to transform complex SPARQL query language to SQL language and conversion algorithm has been proposed to convert value limit pattern.
Key Words：RDF、SPARQL、SQL、Storage Structure、Data Processing
基于关系数据库的RDF数据存储 毕业论文
