零知识下互联数据Web SPARQL查询算法的优化

Zero-knowledge SPARQL query optimization algorithm for the Web of linked data

基于RDF链接遍历的查询方法是一种新颖的对数据Web进行查询的方法,通过在查询执行时不断发现潜在相关的数据源,从而可以充分利用整个数据Web的潜能.结合查询时发现的Descriptor分析原有方案的缺陷,发现原有方案获取了很多对最终结果无贡献的Descroptor,无贡献的Descriptor的获取延长了查询执行时间,降低了查询性能.因此提出一个查询优化算法.通过在查询执行前对查询进行分析,发现对最终结果无贡献的Descriptor,避免从网络上获取,从而减少查询执行时间.实验证明,提出的算法比原有方法拥有更好的查询性能.

RDF link traversal based query execution is a new execution paradigm for the web of data. It allows the execution engine to discover potentially relevant data during the query execution, thus enables users to tap the full potential of the web. In this paper, based on the Descriptor found during the execution of the query, we analyze the restriction of the existing approach and find that the existing approach retrieves many Descriptor that do not contribute to answering the final results The retrieval of these Descriptors extends the query execution time and thus reduces query performance. We present a query optimization algorithm, which analyzes the query before its execution and discovers Descriptors that do not contribute to answering it which can then be prevented from being retrieved, thus shortening the execution time. Compared to the original approach, our algorithm shows that it has a better performance.