摘要
随着社会信息化的高速发展以及大数据时代的到来,科研由技术驱动转向数据驱动,但由于数据量庞大以及数据之间联系的复杂性和隐含性,使用可视化软件构建知识图谱挖掘数据间关联就变得更加迫切。为帮助用户选择合适的可视化软件构建高质量的知识图谱,为用户的合理使用可视化软件提供参考建议。笔者从算法原理的角度,选择可视化过程中的相似度标准化处理、聚类、可视化方法3个主要步骤,对当前两款热门的知识图谱软件Citespace和VOSviewer进行分析,并比较二者功能的优劣差异,研究发现Citespace软件在揭示学科的动态发展规律,发现学科的研究前沿上有一定的优势,而对学科主题之间的关系进行清晰的呈现、或者数据量非常大的时候,可以选取VOSviewer软件绘制知识图谱。
With the rapid development of social informatization and the arrival of the era of big data,scientific research has turned from technology-driven to data-driven.However,due to the large amount of data and the c omplexity and implicitness of the connections between data,it becomes more urgent to use visual software to construct knowledge maps and mine the connections between data.In order to help users choosed the right visualization software to build a high quality knowledge map,and to provide a reference for users to use the visualization software reasonably.This article from the perspective of algorithm principle,choosed the standardization of the similarity in the process of visual processing,clustering and visualization methods of three main steps,analyzed knowl-edge of the current two hot map software Citespace and VOSviewer,and compared the advantages and disadvantages of the two function.The study found that Citespace software has certain advantages in revealing the dynamic development of disciplines and discovering the frontiers of disciplines.When the relationship between subject topics is clearly presented,or when the amount of data is very large,VOSviewer software can be selected to draw knowledge map.
作者
付健
丁敬达
FU Jian;DING Jingda(Department of library and information archives,Shanghai university,Shanghai 200444,China)
出处
《农业图书情报》
2019年第10期31-37,共7页
Agricultural Library and Information
基金
国家社会科学基金项目“数字科研环境下学术信息交流系统的变化规律与模型重构研究”(项目编号:17BTQ071)