以MEDLINE数据库所收录的药理学主题文献集合为例,分别构建全部主题词和主要主题词共现网络,分析了两种网络节点的点度中心度、接近中心度、中介中心度等3种中心性指标. 然后根据节点的度分布,检验了两种网络节点度分布的无标度性,比较分析了两种词共现网络的整体属性(平均距离和聚集系数)以检验其小世界效应. 结果发现,这两种词共现网络都具有无标度特性和小世界效应,由全部主题词构建的共现网络适用于网络属性分析,由主要主题词构建的网络适用于学科主题内容分析.%Co-occurrence network of major MeSH terms and all MeSH terms was constructed respectively with MEDLINE-covered pharmacological literature as its example.The degree, closeness and betweenness of centrality in different nodes of the two co-occurrence networks were analyzed , the distribution of scale-free properties in the two co-occurrence networks was tested, the small-world effect of the two co-occurrence networks was identified by comparing their overall properties ( average distance and clustering coefficient ) , which showed the scale-free properties and small-world effect of the two co-occurrence networks.The co-occurrence network of all MeSh terms could thus be used in analysis of network properties while that of major MeSH terms could thus be used in analysis of subject contents.
展开▼