基于子图的科学家合作网络家族辨识

Family Identification of Cooperative Network of Scientists Based on Subgraph

对科学家合作网络宏观全局结构的研究多表明其小世界和无标度共性特征,而就其微观子图分布特性即网络家族的认知较少。研究基于复杂网络家族辨识的子图比剖面方法、子图组合强度和子图浓度排序方法,系统辨识若干规模和不同领域的科学家合作网络的家族分类及特征。研究表明,具有宏观全局结构共性的上述科学家合作网络,在微观子图的分布、组合或排序上依次具有1种形式、3种规则及5种模式的特性,因而划分为相应家族类别并分别表征理论型、实验型及理论实验型合作网络家族特征。相关方法可用于系统分析和比较科研合作家族行为特性和演化机理。

The research on the macro-global structure of the cooperative network of scientists mostly shows its smallworld and scale-free commonality,while its micro subgraph distribution characteristic,that is,the network family,has less cognition.This paper studies the subgraph ratio profile method based on complex network family identification,the sub-graph combination strength and sub-graph concentration ranking method,and systematically identifies the family classification and characteristics of scientists’cooperative networks of several scales and different fields.The results show that the cooperative networks of scientists have the characteristics of one form,three rules and five patterns in the distribution,combination or ordering of micrographs,which have the commonness of macroscopical and global structure,therefore,it is divided into corresponding family categories and characterizes the family characteristics of theoretical,experimental and theoretical experimental cooperative networks respectively.The related methods can be used to systematically analyze and compare the behavior characteristics and evolution mechanism of cooperative families in scientific research.