语义网站点的发现与排序

Finding and Ranking Semantic Web Sites

随着语义网中RDF数据的大量涌现,语义搜索引擎为用户搜索RDF数据带来了便利.但是,如何自动地发现包含语义网信息资源的站点,并高效地在语义网站点中收集语义网信息资源,一直是语义搜索引擎所面临的问题.首先介绍了语义网站点的链接模型.该模型刻画了语义网站点、语义网信息资源、RDF模型和语义网实体之间的关系.基于该模型讨论了语义网实体的归属问题,并进一步定义了语义网站点的发现规则;另外,从站点链接模型出发,定义了语义网站点依赖图,并给出了对语义网站点进行排序的算法.将相关算法在一个真实的语义搜索引擎中进行了初步测试.实验结果表明,所提出的方法可以有效地发现语义网站点并对站点进行排序.

With the rapid growth of online RDF data, emerging semantic search engines facilitate user＇s searching of RDF （resource description framework） data. It is an open question to all semantic search engines how to find sites containing semantic Web information resources automatically and collect them efficiently. Firstly, the paper introduces a Linking Model of the Semantic Web Sites. The model characterizes the relations among Semantic Web Sites, Semantic Web Information Resources, RDF Models and Semantic Web Entities. This paper discusses the ownerships of Semantic Web Entities based on this model. It also defines a Site Dependency Graph in virtue of the model, and presents a set of ranking algorithms for Semantic Web Sites. Primary tests of these algorithms have been performed in a real-world semantic search engine. Experimental results show that this approach is effective in finding and ranking Semantic Web Sites.