基于科技文献关联数据的科研关系发现研究

发布时间：2018-07-01 11:08

本文选题：科研关系 + 科研实体　；参考：《江苏大学》2017年硕士论文

【摘要】：科技文献作为数据、信息、科研成果的重要载体,不仅包含着大量与学科主题相关的知识体系和知识结构,同时还蕴含着大量的科研实体。不同科研实体之间相互交叉关联将呈现出丰富的科研关系。当前科研关系发现方法主要有:引文分析、共现分析、聚类分析等。这类方法虽然形式化程度较高,但因缺乏本体的支持,无法揭示和识别科研实体之间深层次的语义关系。因此,从语义和本体角度展开对科研关系的关联发现是未来的研究方向和趋势。本研究针对上述问题,结合知识链接、关联数据等方法,展开科研关系关联发现研究。首先,采用文献调研和对比分析的方法,对国内外关联数据、科研关系、知识链接的相关文献进行归纳梳理,在对比分析国内外科研关系发现方法的基础上,提出了利用关联数据进行科研关系发现的思路;其次,采用模型化方法,构建了基于科技文献关联数据的科研关系发现模型,并对模型的功能进行了分析;最后,采用实验的方法对该模型的可用性、易用性、优越性进行验证。通过研究工作的开展,本文取得以下研究成果:(1)提出利用关联数据技术实现科研关系发现的思路,将关联数据技术应用于科研关系发现全过程,丰富了关联数据的应用体系;(2)分析和构建了基于科技文献关联数据的科研关系发现模型,该模型集成整合了现有科研关系发现研究相关的研究成果,充分体现了关联数据在科研关系关联发现方面的优势;(3)借助相关工具验证模型功能,部分实现了科研关系的关联发现,为试图更好的共享、发现、利用科研关系的科研团队和个人,提供一定的帮助和启示。
[Abstract]:As an important carrier of data, information and scientific research achievements, scientific literature not only contains a large number of knowledge systems and knowledge structures related to subject topics, but also contains a large number of scientific research entities. The interrelation between different scientific research entities will present a rich scientific research relationship. At present, the main methods of scientific research relationship discovery are: citation analysis, co-occurrence analysis, cluster analysis and so on. Although this kind of method has a high degree of formalization, it is unable to reveal and identify the deep semantic relationship between scientific research entities because of the lack of ontology support. Therefore, it is the future research direction and trend to develop the association discovery of the scientific research relationship from the semantic and ontological point of view. Aiming at the above problems, the research on association discovery of scientific research relationship is carried out with the methods of knowledge link and association data. First of all, by using the method of literature investigation and comparative analysis, the related data, scientific research relations, knowledge links are summarized and combed, on the basis of comparative analysis of domestic and foreign scientific research relationship discovery methods, This paper puts forward the idea of using the correlation data to discover the scientific research relationship. Secondly, it constructs the scientific research relationship discovery model based on the associated data of scientific literature by using the modeling method, and analyzes the function of the model. The usability, ease of use and superiority of the model are verified by experimental method. Through the development of the research work, this paper has obtained the following research results: (1) put forward the idea of using the correlation data technology to realize the scientific research relationship discovery, and apply the correlation data technology to the whole process of the scientific research relationship discovery. It enriches the application system of correlation data. (2) A scientific research relationship discovery model based on the associated data of scientific literature is analyzed and constructed, which integrates the existing research results of scientific research relationship discovery. Fully reflects the advantages of association data in the scientific research relationship association discovery; (3) with the help of relevant tools to verify the model function, some of the scientific research relationship association discovery, in order to try to better share, discovery, The scientific research team and the individual who use the scientific research relations, provides certain help and enlightenment.
【学位授予单位】：江苏大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：G353.1

【参考文献】