基于云计算的RDF数据存储系统的研究

发布时间：2018-03-02 02:10

本文关键词： 资源描述框架分布式系统存储模型云计算　出处：《合肥工业大学》2013年硕士论文　论文类型：学位论文

【摘要】：语义网的不断发展使得RDF数据存储系统成为一个重要的研究课题。传统RDF数据存储系统在互联网数据爆炸式增长的趋势下遭遇难题，存储能力和查询响应性能都已不能满足用户日益增长的需求。研究人员开始将目光转向云计算领域，云计算具有的海量数据存储和计算能力被认为是解决海量RDF数据存储难题的一个适宜方案，，该方面的研究尚处于起步阶段。本文以海量RDF数据存储系统的设计和实现为目标，对RDF数据存储模型和查询方案进行研究，提出适合云计算环境下的RDF数据存储和查询方案。合理的RDF数据存储模型对于提升RDF数据管理和查询性能具有重要作用。本文在分析现有RDF数据存储模型优缺点的基础上，提出基于有序二叉树的RDF数据存储模型。该模型不仅考虑RDF数据在自身结构方面的属性，而且融入云计算环境下数据存储和处理的特点。该模型在保证RDF数据查询响应效率的同时有效减少了存储开销，且能平滑应对RDF数据快速增长所带来的影响。在RDF数据查询方面，本文通过研究并行计算框架MapReduce的运行机制，结合RDF数据存储模型的特点，以查询性能最优化为目标，提出查询任务生成算法，并证明该算法可有效提升查询响应效率。本文在该算法的基础上，进一步讨论基于有序二叉树的RDF数据模型的特点，给出具体的RDF数据查询响应方案。与传统RDF数据查询方案相比，该方案充分发挥云计算在数据处理方面的优势，有效提高RDF存储系统的数据查询效率。本文最后针对所提出的RDF数据存储模型和查询方案设计了原型系统，并通过实验验证存储模型和查询方案的有效性。
[Abstract]:With the continuous development of semantic Web, RDF data storage system has become an important research topic. Traditional RDF data storage system has encountered difficulties in the trend of explosive growth of Internet data. Storage capacity and query response performance are no longer meeting the growing needs of users. Researchers are turning their eyes to the cloud computing world. Cloud computing has the ability to store and compute massive data, which is considered to be a suitable solution to solve the problem of massive RDF data storage, and the research in this field is still in its infancy. Aiming at the design and implementation of massive RDF data storage system, this paper studies the RDF data storage model and query scheme, and proposes a RDF data storage and query scheme suitable for cloud computing environment. Reasonable RDF data storage model plays an important role in improving RDF data management and query performance. This paper analyzes the advantages and disadvantages of existing RDF data storage models. This paper presents a RDF data storage model based on ordered binary tree, which not only considers the properties of RDF data in its own structure. This model not only ensures the efficiency of RDF data query response, but also reduces the storage cost, and can smoothly cope with the impact of the rapid growth of RDF data. In the aspect of RDF data query, by studying the running mechanism of parallel computing framework MapReduce, combining the characteristics of RDF data storage model and aiming at the optimization of query performance, a query task generation algorithm is proposed in this paper. It is proved that the algorithm can effectively improve the query response efficiency. Based on the algorithm, this paper further discusses the characteristics of RDF data model based on ordered binary tree. Compared with the traditional RDF data query scheme, the scheme gives full play to the advantages of cloud computing in data processing, and effectively improves the efficiency of data query in RDF storage system. Finally, a prototype system is designed for the proposed RDF data storage model and query scheme, and the validity of the storage model and query scheme is verified by experiments.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP333

【相似文献】