引文内容视角下的引文网络知识流量研究  

Research on Citation Network Knowledge Flow from the Perspective of Citation Content

作　　者：叶光辉[1] 彭泽 毕崇武 李明倩 Ye Guanghui

机构地区：[1]华中师范大学信息管理学院,湖北武汉430079

出　　处：《情报理论与实践》2020年第12期11-18,共8页Information Studies:Theory & Application

基　　金：国家自然科学基金青年项目“基于标签语义挖掘的城市画像计算与应用模型研究”(项目编号:71804055);中央高校基本科研业务费资助项目“基于引文分析的跨地域知识流动效率及其影响因素研究”(项目编号:2019CXZZ126)的成果之一。

摘　　要：[目的/意义]为了弥补基于题录数据的引文网络分析法忽略引用行为中语义信息的问题,文章尝试从引文内容视角探究知识在学术文献间的流动状况。[方法/过程]首先,提取文献知识主题,用知识描述规则判断知识主题所在的文本片段是否含有知识;其次,将知识主题对应的施引与被引文本的相似度作为该知识的被引程度;最后,结合被引文献知识存量得到引文网络中节点间的知识流量。[结果/结论]文章从CNKI数据库中获取1690篇"智库"主题文献,按照上述方法依次提取文献知识主题、量化节点知识存量、计算节点间知识被引程度,最终构建了基于引文内容的知识流动网络,充分发掘了引用行为中有关知识流动的信息,很好地展示了知识在引文网络中的流动情况。[Purpose/significance]In order to make up for the problem that citation analysis based on bibliographic data ignores semantic information in citation behavior,this paper attempts to explore the flow of knowledge among literatures from the perspective of citation content.[Method/process]Firstly,the topic of literature knowledge is extracted and the rules of knowledge description are used to judge whether the text fragment where the topic of knowledge is located contains knowledge.Then the similarity between the citing text and the cited text mapped by the knowledge topic is calculated as the degree of citation.Finally,the knowledge flow between nodes in the citation network is obtained based on the knowledge stock of cited references.[Result/conclusion]This paper obtained 1690"think tank"themed literatures from CNKI database.After extracting the knowledge topic of the article,quantifying the knowledge stock of nodes,and calculating the flow of knowledge flow between nodes,the knowledge flow network was finally constructed and analyzed.The information about knowledge flow in citation behavior is fully explored and the flow of knowledge in citation network is well demonstrated.

关 键 词：引文内容 引文网络 知识流动 文本相似度 智库 文献计量 

分 类 号：G353.1[文化科学—情报学]