摘要
[目的/意义]为改变以往引文网络主路径相关研究将所有学术文献及其之间引用关系一视同仁而忽略连边语义信息的做法,文章设计一种引文内容视角下的知识流动主路径提取方法。[方法/过程]首先通过一种引文内容视角下的知识流量计算方法计算节点之间的知识流量;然后从引用行为的情感倾向和发生位置两个维度来判断节点间知识流动路径的类型;再通过搜索路径统计数(SPC)算法计算网络中各节点间连线被网络中所有路径遍历的次数;最后融合节点间知识流量、知识流动路径类型以及搜索路径统计数等因素来衡量该连边在网络中重要性,并基于此提取出知识流动主路径。[结果/结论]以"数字人文""数据治理"和"智库"三个主题为例,从中国知网分别获得1192、2136和6566篇文章,按照上述方法构造知识流动网络并提取出主路径。经过分析,得到的知识流动主路径可以很好地反映知识在引文网络中的客观流动情况,对挖掘学科内知识交流热点、促进学科内及学科间知识交流行为具有指导意义。
[Purpose/significance]Previous studies on the main path of citation networks treated all academic literatures and their citation relations alike and ignored the link semantic information.In order to change this practice,this paper designed a method to extract the main path of knowledge flow from the perspective of citation content.[Method/process]Firstly,a method of calculating the mount of knowledge flow from the perspective of citation content is used to calculate the knowledge flow between nodes;Then the types of knowledge flow path between nodes are judged from the emotional tendency of reference behavior and the location of reference behavior;Then,the SPC algorithm is used to calculate the number of times that the links between nodes in the network are traversed by all the paths in the network;Finally,factors such as the mount of knowledge flow between nodes,the type of knowledge flow path and the value of SPC are integrated to measure the importance of this connection in the network,and the main path of knowledge flow is extracted based on this.[Result/conclusion]Taking"digital humanity","data governance"and"think-tank"as examples,this paper obtains 1192,2136 and 6656 articles from CNKI respectively,constructs the knowledge flow network according to the above methods,and extracts its main path of knowledge flow.After analysis,the main path of knowledge flow obtained in this paper can well reflect the objective flow of knowledge in the citation network,which is of guiding significance to excavate hot spots of knowledge exchange within disciplines and promote knowledge exchange within and between disciplines.
出处
《情报理论与实践》
CSSCI
北大核心
2020年第12期19-25,10,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。