【作者】 陆铭彦；

【导师】 吴慧中；

【作者基本信息】 南京理工大学 ， 计算机应用技术， 2007， 硕士

【摘要】 随着网络技术的高速发展，分布式已成为当前信息处理、传播与交换的重要的网络体系结构。各种数据资源在不同地点、以不同方式存储在分布异构的数据库中，由此导致了分布式网络中数据资源的共享、集成等问题。为了更好地解决数据资源的分布性与异构性问题，在对现有异构数据集成系统和各类技术分析比较的基础上，提出了一个基于网格技术与RDF技术的数据查询集成方案。方案采用网格技术来解决数据源的分布性和访问异构性，并采用RDF技术来解决数据源间的语法异构和语义异构性。利用此方案，构建了基于RDF的网格数据查询集成系统(RGDS)，分析了该系统中各主要模块的功能及它们之间的联系。在网格数据集成系统中，数据源质量千差万别，存在数据源的选择问题。为此，定义了数据源相关信任度的概念，以此来衡量数据源服务的质量，给出了数据源信任度的计算方法，提出了基于信任度的数据源贪婪选择算法，并通过实验验证了该方法在数据源选择上的有效性。最后介绍了基于RDF的网格数据查询集成系统(RGDS)的实现，包括网格数据源服务的注册与管理，并通过一个实例详细阐述了RGDS系统的流程与各功能模块的具体实现，简单介绍了基于RDF的数据查询结果的推理方法。更多还原

【Abstract】 With the development of the network technology, the distributed structure has beenthe important network architecture for information processing、transmitting andexchanging. All kinds of data sources are stored in the distributed and heterogeneousdatabases in different locations and with different forms. So this situation leads to theproblems of data sharing and integration in the distributed network.The data integration systems and technologies in existence are introduced. In orderto solve the distributed and heterogeneous problems of the data sources, a new solutionfor data integration based on grid services and RDF technology is given. Grid servicesare used to solve the heterogeneous problem of data accessing, and RDF technology isused to solve the heterogeneous problems of grammar and semantics. According to thesolution, we establish a RDF-Based Grid-Data Query and Integration System (RGDS), and analyze the main modules’ functions and their relationships.Because of the differences of data source quality, it is a problem that how to select thecorrect and appropriate data source in the data query system. So we define the trust valueto evaluate the quality of data source. The greedy-selecting algorithm based on trustvalues of data sources is proposed. A preliminary experiment is carried out to evaluatethe efficiency of the proposed algorithm.Finally, the realization procedures of RDF-Based Grid-Data Query and IntegrationSystem (RGDS) are illustrated by an instance. Also, the reasoning of RDF data results isintroduced simply.更多还原