图书情报工作 ›› 2016, Vol. 60 ›› Issue (10): 115-122.DOI: 10.13266/j.issn.0252-3116.2016.10.016

• 知识组织 • 上一篇    下一篇

基于期刊主题相似性的领域分析数据集构建:方法与实证

刘敏娟, 张学福, 颜蕴, 陈露   

  1. 中国农业科学院农业信息研究所 北京 100081
  • 收稿日期:2016-05-03 出版日期:2016-05-20 发布日期:2016-05-20
  • 通讯作者: 张学福,知识工程研究室主任,研究员,博士,通讯作者,E-mail:zhangxuefu@caas.cn
  • 作者简介:刘敏娟(ORCID:0000-0001-8422-2919),馆员,博士研究生;颜蕴,文献资源发展部主任;陈露,硕士研究生

Method and Empirical Study on Data Set Construction in Domain Analysis Based on Journal Topic Similarity

Liu Minjuan, Zhang Xuefu, Yan Yun, Chen Lu   

  1. Agricultural Information Institute of Chinese Academy of Agricultural Sciences, Beijing 100081
  • Received:2016-05-03 Online:2016-05-20 Published:2016-05-20

摘要: [目的/意义] 重点针对一些在数据库中无法找到既定的主题分类,无法罗列完整关键词,也不可能通过选择有代表性机构和现成的学科领域核心期刊的方法实现数据集构建的领域,提出一种基于期刊主题相似性的领域分析数据集构建的方法。[方法/过程] 该方法组合运用引文分析与期刊文献耦合分析方法,并借助科学知识图谱绘制方法,通过确定学科领域的代表性期刊群组,经过不同形式的组配最终达到满足不同层次需求的构建数据集的目的。[结果/结论] 本方法可以满足宏观、中观和微观不同领域分析层次的需求,操作过程简单灵活且人工干预的程度不高,通过在具体领域的实例验证,证明其可以有效地解决一些领域数据集构建的难题,对今后相关研究具有一定借鉴意义。

关键词: 数据集构建, 期刊主题相似性, 引文分析, 期刊文献, 耦合分析, 知识图谱

Abstract: [Purpose/significance] This article proposes a method of data set construction in domain analysis based on journal topic similarity, which focuses on solving the problem that data set construction cannot be realized in the cases that established subject classification cannot be gained and complete keywords be listed in some databases, and typical institutions and existing core journals in a filed cannot be chosen.[Method/process] This method combines the citation analysis, journal literatures couple analysis and mapping knowledge domains to determine the representative journal groups of a subject, and then constructs the data set to meet the different demands by different combination.[Result/conclusion] This method can meet the demand of different levels of macro, medium and micro, with simple and flexible process but a low degree of human intervention. It is proved that it can effectively solves the problem of data sets construction in some field by empirical study of a specific field, and has reference value for the related research in future.

Key words: data set construction, journal topic similarity, citation analysis, journal literatures, couple analysis, knowledge map

中图分类号: