基于SPARQL的链接流数据查询方法研究

发布时间：2018-04-27 08:03

本文选题：语义物联网 + 链接数据　；参考：《大连海事大学》2017年硕士论文

【摘要】：本文分析了互联网中数据的特性,如实时性、海量性、异构性等。研究数据的识别与访问、海量数据查询、实时查询等问题。随着语义技术的发展,语义网与物联网的结合,出现了链接流数据。链接流数据解决了数据的识别和访问问题,但现有链接流数据查询处理系统有着各种各样的不足。为了更好地处理链接流数据,本文对链接流数据查询方法进行了研究。首先,本文详细介绍了现有链接流数据查询系统,通过实验进行比较分析,指出了现有查询系统的不足;其次,本文提出了面向链接数据的优化方法和基于增量的运算符结构。在链接流数据处理过程中,静态的链接数据会影响处理速度,为了解决这个问题,本文对RDF图进行了分析,提出了一种面向链接数据的优化方法。该方法通过对RDF图进行压缩,提高了在RDF图上的数据处理速度。在链接流数据查询过程中,主要使用运算符进行实现,本文分析了现有系统中运算符的功能,设计了基于增量的运算符结构。这种结构可以使运算符更好地适应连续处理过程,减少了链接流数据查询所需要的时间;最后,本文设计了基于SPARQL的链接流数据查询系统并加以实现。本文将提出的链接流数据查询方法与现有的查询方法进行了对比分析,结果表明,本文所提出的方法提高了链接流数据的查询速度。
[Abstract]:This paper analyzes the characteristics of data in the Internet, such as real-time, magnanimity, heterogeneity and so on. Research data identification and access, massive data query, real-time query and other issues. With the development of semantic technology and the combination of semantic Web and Internet of things, link flow data appear. Link flow data solves the problem of data identification and access, but the existing link flow data query processing system has various shortcomings. In order to deal with link stream data better, this paper studies the query method of link stream data. Firstly, this paper introduces the existing link stream data query system in detail, and points out the deficiency of the existing query system through the comparison and analysis of experiments. Secondly, this paper proposes the optimization method for link data and the operator structure based on increment. In the process of link flow data processing, static link data will affect the processing speed. In order to solve this problem, this paper analyzes the RDF diagram and proposes an optimization method for link data. By compressing the RDF diagram, the data processing speed on the RDF diagram is improved. In the process of link stream data query, operators are mainly used to implement. This paper analyzes the functions of operators in existing systems, and designs the structure of operators based on increment. This structure can make operators better adapt to the continuous processing process and reduce the time required for link stream data query. Finally, a link stream data query system based on SPARQL is designed and implemented in this paper. In this paper, the link stream data query method is compared with the existing query methods. The results show that the proposed method improves the query speed of link stream data.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.3

【参考文献】