中国大学关系的复杂网络

采用14个参量描述大学状态,这些状态间的余弦距离定量描述各大学之间的相似性,得到最小生成树.每个大学与最相似的若干大学连接,得到描述大学关系的复杂网络.度、介数、紧密度和社区结果等拓扑概念用于讨论这些网络的非平凡结构特点.中国大学(不含港澳台地区院校)划分为12个社区,每个社区内度最大的节点作为社区标志,而社区间邻接的介数最大的点作为边界.发现边界是排名在前列的具有强竞争力的高校,表明"综合化"是大学提高竞争力的努力方向.紧密度小的边界和各社区标志大学可作为中国大学代表,可供综合评价大学教育体系参考.

Complex network approach to relations between universities in China

Totally 14 variables are used to describe the state of each university.Properties such as degree,betweenness,closeness,and community are employed to detect nontrivial characteristics in the cosine distance-based spanning tree and the university relation network.It is found that the relation network can be divided into 12 communities,namely,the considered universities can be separated into 12 groups.Each group is represented with the center node,namely,the intra-node with the largest degree,and the boundary node,namely,the node with the largest value of betweenness among all the nodes contact with other communities.A set of universities,including the representatives of the communities and the boundary nodes with smaller values of closeness,is proposed to be reference universities in globally evaluation of the high-education system.