基于关联数据的知识元链接图式存储模型研究

发布时间：2018-03-01 11:17

本文关键词： 关联数据知识元链接图式存储知识元抽取　出处：《华中师范大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着信息技术和网络技术的迅速发展,用户的知识信息需求发生了深刻变革。从用户认知的角度来看,通过链接方式获取所需信息,符合用户通过联想方式获取信息的本质,因此,以知识元链接为核心的知识服务逐步成为用户获取知识信息的重要手段。目前,以关系-实体模型对知识元链接数据进行的存储存在一些不足,首先,数据表中数据之间的联系性很难体现出来,这是由E-R模型本身的特点所决定的；其次,E-R模型只强调数据的静态组织,数据与其操作是分离的,缺少对数据完整性描述；再次,E-R模型无法对实体之间的各种联系进行进一步的约束,最后,查询语言不能有效地挖掘数据之间存在的语义链接关系。以上E-R模型存在的缺点和不足将严重影响数据存储效率。近年来,随着关联数据技术研究的不断深入和成熟,使用关联数据技术对知识元链接数据进行语义表示,进而为实现机器可读的、链接更丰富的知识元提供了有效的途径和方法。文章结合关联数据发布的四项基本原则,通过RDF数据模型,将非结构化的、异构的知识元对象按照统一的标准进行一致性转化,最终将这些数据以原有的知识链接图的形式存储于图数据库中。本文提出的基于关联数据的知识元链接图式存储模型研究是对传统知识元链接存储模式的一种改进,能够有效实现关联数据环境下知识元链接的图式存储。本文以关联数据为视角,对知识元链接的图式存储模型构建理论与方法进行了探讨。首先对知识元链接、知识元抽取、图式存储和关联数据相关概念及理论进行综述,然后探讨知识元结构的分析与确定、知识元抽取的方法、知识元标引及关联数据在知识元链接图式存储中的作用。其次,在分析传统知识元链接存储模式存在不足的基础上,提出关联数据环境下知识元链接图式存储模型,并从知识元抽取模块、知识元链接语义描述模块、语义知识元链接存储模块和语义知识元链接图式存储应用模块四个方面对该模型进行详细阐述。最后,以科技文献知识元为实验数据来源,对该模型进行实验验证,论证该模型的可行性。
[Abstract]:With the rapid development of information technology and network technology, profound changes have taken place in the knowledge and information needs of users. Users from a cognitive point of view, to obtain the required information through the link, consistent with the nature, access to information users through the associative way therefore, knowledge service knowledge element link as the core has gradually become an important means to acquire knowledge and information users at present, the knowledge element linking data to relational entity model storage has some shortcomings, first of all, the connection between the data in the table is difficult to be reflected, which is decided by the characteristics of E-R model by E-R model; secondly, emphasize the static data organization, data and operation isolation, lack of data integrity; thirdly, E-R model to various links between entities are further constrained, finally, query language can not effectively tap number The shortcomings and shortcomings of the above E-R model will seriously affect the efficiency of data storage.
In recent years, with the deepening of research on data association and mature, using the related data of knowledge element link data semantic representation, then in order to realize the machine readable, knowledge element link provides more effective ways and methods. According to the four basic principles of linked data released by RDF, data model will, unstructured, heterogeneous knowledge element object in accordance with the unified standard consistency transformation, finally these data will be stored in the form of original knowledge link diagram in graph databases. In this paper, the research of related data storage model based on schema knowledge element link is an improvement to the traditional storage mode of knowledge element link the association can effectively achieve the data environment of knowledge element link storage schema.
Based on the related data from the perspective of schema storage model of knowledge element link construction theory and method are discussed. Firstly, knowledge element link, knowledge element extraction, review the schema storage and associated data related concepts and theories, then discusses the analysis and determination of the structure of knowledge element, knowledge element extraction method, the role of knowledge element the indexing and the associated data in the gragh storage of knowledge element link. Secondly, based on the analysis of the traditional storage mode of knowledge element link problems, put forward gragh storage model of knowledge element link correlation data environment, and from the knowledge element extraction module, knowledge element linking semantic description module, the four aspects of the semantic knowledge element link storage module and the semantic knowledge element linking schema storage application module in detail of the model. Finally, the science and technology literature knowledge element as the source of the experimental data, the model is real Verify the feasibility of the model.

【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP311.13;TP333

【参考文献】