领域演化结构识别的主路径方法与高被引论文方法对比研究  

Comparative Study of Main Path Analysis and Highly-cited-paper Clustering to Identify Domain Evolving Structure

作　　者：韩毅[1] 童迎[1] 夏慧[1] 

机构地区：[1]西南大学计算机与信息科学学院

出　　处：《图书情报工作》2013年第3期11-16,共6页Library and Information Service

基　　金：<图书情报工作>杂志社出版基金项目"基于引文网络主路径的领域演化结构识别研究"(项目编号:2012CB002);西南大学科技基金资助项目"基于引文网络主路径的学科演化结构识别与演化动力学机制研究"(项目编号:SWU112001)研究成果之一

摘　　要：高被引论文方法通过节点相似性聚类来反映领域演化结构,而主路径方法则通过节点连通性来揭示领域演化结构。在分析两种方法差异的基础上,以富勒烯领域引文数据为样本,应用HistciteTM和Pajek对比研究主路径组分完全网络和相同数量顶点的高被引论文网络,结果表明二者的结构与组分差异显著。主路径组分的当年引用数排序的结果表明,多数路径组分不完全是当年的高被引论文,进而证明了主路径分析方法的独特性。Highly cited papers method is commonly used to reflect the domain evolving structure by calculating the node similarity, while the main path analysis reflects the domain evolving structure by calculating the node connectivity. Based on the differences between these two methods and taking fullerene citation data as the sample, this paper comparatively studies the complete network of main path component and highly-cited-paper network of same number vertex with Pajek and HistCiteTM. The result shows that the network structure and components are significantly different between of them. Furthermore, the cited counts of the main path node every year reveal that most path components are not completely the highly-cited papers and that the method of main path analysis is unique.

关 键 词：主路径分析 高被引论文聚类 领域演化结构 HistCiteTM PAJEK 

分 类 号：G353.1[文化科学—情报学]