主题-主题关联的学科知识网络构建与演化分析  

Research on Construction and Evolution Analysis of Discipline Knowledge Network Based on Topics Association

作　　者：王曰芬[1] 王金树 关鹏[1,2] WANG Yue-fen;WANG Jin-shu;GUAN Peng(School of Economics and Management,Nanjing University of Science ＆ Technology,Nanjing 210094,China;Institute of Applied Mathematics,Chaohu University,Hefei 238000,China)

机构地区：[1]南京理工大学经济管理学院,江苏南京210094 [2]巢湖学院应用数学学院,安徽合肥238000

出　　处：《情报科学》2018年第9期9-15,102,共8页Information Science

基　　金：国家自然科学基金(71373124)

摘　　要：【目的/意义】构建主题-主题关联的学科知识网络,从主题之间语义关联角度度量主题在学科知识网络中的影响力,分析学科知识结构演化规律,为热点、前沿主题探测提供方法支撑。【方法/过程】基于LDA主题模型抽取学科领域研究主题,利用主题在科学文献中的共现关系构建主题-主题关联的学科知识网络,并提出主题影响力概念和度量方法;基于复杂网络结构分析方法对学科领域生命周期内主题-主题关联的学科知识网络进行演化分析。【结果/结论】实证分析表明主题的网络影响力是主题强度、被学者关注度等外部特征指标的有力补充,可用来探测热点、前沿主题。同时,学科知识结构随着学科领域的发展表现出较强的小世界网络特征。【Purpose/significance】The purpose of this paper is constructing and analyzing the discipline knowledge network based on topics association. Meanwhile, the influence of topics could be measured through semantic association of topics, in order to find hot and frontier research topic.【Method/process】To construct the discipline knowledge network based on topics association, the paper first extracts discipline research topics by topic model, then computes co-occurrence intensity between topics. Meanwhile, we put forward the concept of topic influence and propose the method of measuring topic influence. Next, the paper puts forward evaluation analysis method of network structure, by complexity network structure analysis theory.【Result/conclusion】We get two important results through empirical analysis. Firstly, the network influence of topic is a powerful supplement to topic intensity and scholars＇ attention. So, using this method, we can find hot and frontier research topics in discipline. Secondly, the discipline knowledge network based on topics association shows stronger smallworld network characteristics with the development of discipline area.

关 键 词：学科知识网络 复杂网络 演化分析 主题影响力 网络科学 

分 类 号：G201[文化科学—传播学]