领域本体RDF图数据存储与查询方法研究

发布时间：2018-10-21 17:34

【摘要】：领域本体是指对特定领域内概念及概念间关系的形式化表达,本质上是一种共享的资源描述机制,用于描述一个领域内的知识信息,通常采用OWL、RDF等语言描述。领域本体一般采用OWL文件或关系型数据库来存储。随着信息获取、存储和传播技术的飞速发展,民航突发事件应急管理领域本体面临复杂图数据处理的新问题。关系型数据库与本体数据复杂的图形结构在模型上的不匹配从根本上制约了大规模本体数据的处理能力。Neo4j是当前比较成熟且高性能的开源图数据库,遵循属性图数据模型,其完全支持ACID事务、支持海量数据及迅速的图查询等特点,可以较好地解决领域本体复杂的图结构问题。为此,本文将Neo4j图数据库存储技术方法与民航突发事件应急管理领域本体相结合,进行了以下的研究:(1)在分析民航突发事件应急管理领域本体RDF数据存储与应用现状的基础上,提出了一种基于Neo4j的领域本体RDF图数据存储方法,通过领域本体RDF有向标记图与Neo4j图数据库映射模型的构建,给出了RDF图与Neo4j图之间的映射规则定义,在此基础上,采用与MySQL存储方案相比较的方法,进一步验证了该存储方案的高效性,解决了民航突发事件应急管理领域本体RDF图数据的存储与查询效率低的问题。(2)针对Neo4j图数据库不支持分布式处理的问题,设计了一个逻辑处理层来负责分布式相关的逻辑处理以支持Neo4j的领域本体分布式存储,给出了分布式扩展的总体存储框架,并对该存储框架中主要模块的设计细节与相关算法进行了阐述,采用基于动态图数据流的划分方法,实现了民航突发事件应急管理领域本体RDF图的动态划分,并通过实验验证了该分布式存储的特性,为基于图数据的民航突发事件应急管理领域本体的大数据应用提供了方法支撑。
[Abstract]:Domain ontology is a formal representation of concepts and their relationships within a particular domain. In essence, it is a shared resource description mechanism, which is used to describe knowledge information in a domain, usually in OWL,RDF and other languages. Domain ontology is generally stored in OWL files or relational databases. With the rapid development of information acquisition, storage and dissemination technology, the field of emergency management of civil aviation emergencies faces a new problem of complex graph data processing. The mismatch between the complex graph structure of relational database and ontology data fundamentally restricts the processing ability of large-scale ontology data. Neo4j is a mature and high performance open source graph database which follows the attribute graph data model. It fully supports ACID transactions, supports massive data and quick graph query, and can solve the complex graph structure problem of domain ontology. Therefore, this paper combines Neo4j map database storage method with civil aviation emergency management domain ontology, and carries out the following research: (1) on the basis of analyzing the current situation of RDF data storage and application in civil aviation emergency management domain, A data storage method of domain ontology RDF graph based on Neo4j is proposed. The mapping rules between RDF graph and Neo4j graph are defined by constructing the mapping model of domain ontology RDF directed label graph and Neo4j graph database. Compared with the MySQL storage scheme, the efficiency of the storage scheme is further verified. The problem of storage and query efficiency of ontology RDF graph data in the field of civil aviation emergency management is solved. (2) the Neo4j graph database does not support distributed processing. In this paper, a logical processing layer is designed to be responsible for distributed logic processing to support the distributed storage of domain ontology in Neo4j, and the general storage framework of distributed extension is given. The design details and related algorithms of the main modules in the storage framework are described. The dynamic partition of the ontology RDF diagram in the field of emergency management of civil aviation emergencies is realized by using the partition method based on the data flow of the dynamic graph. The characteristic of distributed storage is verified by experiments, which provides the method support for big data application of civil aviation emergency management field based on graph data.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.13

【参考文献】