摘要
【目的/意义】共词分析法是文献情报分析的重要方法,对促进学科领域等的研究和发展具有重要作用。针对共词分析法中存在的文献层面和词层面"同量不同质"问题、高频孤立词问题等,提出一种基于文献多属性加权的共词分析方法。【方法/过程】该方法利用文献发表时间、被引次数和下载量等属性来度量文献时效性和被关注度,以表征文献层面的权重;利用文献标题和摘要等属性对文献关键词的差异性进行度量;综合文献权重和词差异性权重对共词分析词频次及共词对频次进行加权处理,然后依据二者加权后的频次确定共词分析用词集合及构建共词矩阵。【结果/结论】以CNKI数据库有关主题模型研究的核心期刊论文为数据源,对该领域的研究主题热点进行分析,从共词分析用词选择和聚类分析等方面与一般共词分析法进行对比,验证了本文所提加权方法具有一定的合理性和有效性。
【Purpose/significance]Co-word analysis is an important method for literature information research,which is of great significance to discipline development.However,the method has the problem of different quality with the same quantity both in document level and word level and the problem of high-frequency isolated word.【Method/process]This paper proposes a weighted co-word analysis method based on multi attributes of literature,which uses literature publication time to measure timeliness and uses literature citation times and downloads to measure attention degree,and then uses the two indicators to measure the weight of a literature.Moreover,the importance of each word is measured by the literature’s title and abstract.Then,take the literature weight and word importance to improve the general co-word analysis method by weighted co-word frequency calculation and chosen co-word matrix construction.【Result/conclusion]In case of Topic Model research,this paper analyzes the domain research topic and hotspot based on the data collected from CNKI database and the result shows,compared with general co-word analysis method,the improved method is proved to be some reasonable and effective at word choice and co-word clustering analysis.
作者
余本功
王龙飞
陈杨楠
杨颖
YU Ben-gong;WANG Long-fei;CHEN Yang-nan;YANG Ying(School of Management,Hefei University of Technology,Hefei 230009,China;Key Laboratory of Process Optimization&Intelligent Decision-making,Ministry of Education,Hefei University of Technology,Hefei 230009,China)
出处
《情报科学》
CSSCI
北大核心
2019年第1期122-128,共7页
Information Science
基金
国家自然科学基金项目:“基于制造大数据的产品研发知识集成与服务机制研究”(71671057)、“不确定环境下的复杂产品研发协同绩效动态评价研究”(71573071)
过程优化与智能决策教育部重点实验室开放课题资助
关键词
加权共词分析
文献属性
时效性
被关注度
词差异性
weighted co-word analysis
literature attribute
timeliness
attention degree
word difference