多维度视角下学科主题演化可视化分析方法研究——以我国图书情报领域大数据研究为例
发布时间：2019-02-19 17:56
【摘要】：探测、识别某学科领域研究主题的演化过程并进行可视化分析,对于掌握研究现状和发展趋势具有重要意义。学科主题演化是一个复杂过程,存在多种变量,如主题强度、结构和内容等,目前研究主要以单一维度进行可视化分析,信息负荷过大,存在感知局限性。本文提出多维度视角下学科主题演化可视化分析方法:通过人工标注方法对关键词进行语义角色分类,利用Fast Unfolding算法识别出具有语义特征的学科主题;利用余弦相似度计算公式计算学科主题相似度判定演化关系;构建多维度学科主题演化分析模型,并设计了三种创新性的科学知识图谱,进行学科主题强度、结构和内容三个维度的可视化分析,通过相互作用可以帮助快速消化、理解信息和精炼分析结果,有效地分析学科主题演化的复杂过程。通过对我国图书情报领域近10年大数据研究的实证分析,证明该方法具有可行性和有效性。
[Abstract]:It is of great significance to detect, identify and visualize the evolution of research topics in a certain discipline. Subject evolution is a complex process with many variables such as theme intensity structure and content. At present the research mainly uses a single dimension to visualize the analysis. The information load is too heavy and there are perceptual limitations. In this paper, a visual analysis method of subject evolution from multi-dimensional perspective is proposed: the semantic role classification of keywords is carried out by manual annotation method, and the subject with semantic characteristics is recognized by Fast Unfolding algorithm. Using cosine similarity formula to calculate the subject similarity to determine the evolutionary relationship; A multi-dimensional thematic evolution analysis model is constructed, and three innovative maps of scientific knowledge are designed to analyze the intensity, structure and content of the subject, which can help the rapid digestion through the interaction. Understand the information and refine the analytical results, effectively analyze the complex process of subject evolution. This method is proved to be feasible and effective by empirical analysis of big data in the field of library and information in China in recent 10 years.
【作者单位】： 山东理工大学科技信息研究所;
【基金】：国家社会科学基金项目“未来新兴科学研究前沿识别研究”(编号:16BTQ083)的研究成果之一~~
【分类号】：TP399-C1;G353.1
本文编号：2426726
[Abstract]:It is of great significance to detect, identify and visualize the evolution of research topics in a certain discipline. Subject evolution is a complex process with many variables such as theme intensity structure and content. At present the research mainly uses a single dimension to visualize the analysis. The information load is too heavy and there are perceptual limitations. In this paper, a visual analysis method of subject evolution from multi-dimensional perspective is proposed: the semantic role classification of keywords is carried out by manual annotation method, and the subject with semantic characteristics is recognized by Fast Unfolding algorithm. Using cosine similarity formula to calculate the subject similarity to determine the evolutionary relationship; A multi-dimensional thematic evolution analysis model is constructed, and three innovative maps of scientific knowledge are designed to analyze the intensity, structure and content of the subject, which can help the rapid digestion through the interaction. Understand the information and refine the analytical results, effectively analyze the complex process of subject evolution. This method is proved to be feasible and effective by empirical analysis of big data in the field of library and information in China in recent 10 years.
【作者单位】： 山东理工大学科技信息研究所;
【基金】：国家社会科学基金项目“未来新兴科学研究前沿识别研究”(编号:16BTQ083)的研究成果之一~~
【分类号】：TP399-C1;G353.1
【相似文献】
相关期刊论文 前9条
1 王春秀;冉美丽;;学科主题演化定量分析的理论基础探析[J];现代情报;2008年06期
2 杜育敏;;“多学科主题整合”课程改变教与学的方式[J];现代教学;2014年Z3期
3 史艳丽;张翠玲;;网络学科主题门户研究概述[J];农业图书情报学刊;2006年07期
4 朱梦娴;程齐凯;陆伟;;基于社会网络的学科主题聚类研究[J];情报杂志;2012年11期
5 赵凡;;基于共词分析的学科主题动态跟踪相似算法改进研究[J];情报杂志;2010年01期
6 庞杰;刘则渊;姜春林;;基于学科主题共现的国际能源期刊分类与知识结构研究[J];科学学与科学技术管理;2010年09期
7 赵凡;;基于共词分析的学科主题动态跟踪相似方法探讨[J];情报杂志;2009年09期
8 卿蔚;雷菊霞;吴英梅;;国外学科主题网关的发展及其启示——以Intute:social science为例[J];情报资料工作;2008年06期
9 ;[J];;年期
相关会议论文 前1条
1 黎彬;阮学平;任慧玲;葛红梅;孙晓北;;略论医学图书馆文献资源学科配置的合理化[A];中国图书馆学会专业图书馆分会2009年学术年会论文集[C];2009年
相关硕士学位论文 前1条
1 汪雅喜;基于共词网络的学科主题结构与演化研究[D];浙江理工大学;2013年
，本文编号：2426726
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2426726.html
