本文通过对2005年北京师范大学图书借阅记录的分析,构建了单顶点网和二分网两种网络,给出了图书借阅数据在这两个网络上的度分布,通过比较得到这两个网络的度分布具有时间演化不变性。然后在单顶点网络上抽取了部分数据进行分析,发现各子网络的平均路径长度L和网络的平均集聚系数C随网络规模增大总体上有减小的趋势;并且网络具有相对较高的平均集聚系数、较小的“平均最短路径长度”,表现出明显的“小世界效应”。最后在二分网上基于同类节点的相似性提出了基于三元组的集聚系数的定义,给出了该集聚系数在北师大图书借阅网络上的分布.
By analyzing borrowing and reading record of books in library of Beijing Normal University during 2005, we construct two kinds of networks--unipartite networks and bipartite networks, then we obtain the degree distribution of data of berrowing and reading record of books on these two networks. By comparison the invariability with time of degree distribution has been found out. Then part of data from nnipartite networks has been analyzed and the property that the average path length and average clustering coefficient of sub-networks have inverse relationship with size of whole network has been found out. Moreover, the network has higher average clustering coefficient and smaller "the shortest average path length", which shows obvious character of "small world". Finally definition of the cluster coefficient based on connected triples of bipartite networks has been brought out and the distribution of the cluster coefficient in library networks of Beijing Normal University has been got.