社交网络拓扑结构实证研究——以科学网博客为例

An Empirical Study on Topological Structure of Online Social Network——Take Science Net Blog as example

采用传统统计方法很难直观了解大规模社交网络的结构特点和演化特征。通过对科学网博客域名下的网页进行搜索,建立一个由244 662个博主和113 062对好友关系构成的复杂网络——科学网博客博主好友关系网络。采用复杂网络理论进行研究,测算网络度分布、平均路径长度和聚类系数,发现该网络具有无尺度属性和小世界属性,存在相对较多的集散节点,导致网络的度分布幂指数小于正常范围。通过逐步删除高连接度节点,观察网络破碎程度,分析了集散节点在维持社交网络链接中的重要性,建议重点关注10%的最高度节点,使网络更加健壮。该研究有助于阐明在线社交网络的自组织结构性质。

It is difficult to intuitively understand the structural characteristics and evolution characteristics of large-scale social network by using traditional statistical methods.By searching the web pages under the science website,we have established a complex network composed of 244 662 bloggers and 113 062 friends'-sciencenet blog bloggers friends relationship network.The complex network theory is used to study the network,and the network degree distribution,the average path length and the clustering coefficient are measured.It is found that the network has free scale attribute and small world attribute.It is analyzed that the exponential of the degree distribution of the network is less than the normal range due to the existence of relatively large hubs in the network.Finally,we study the size of the giant component as gradually deleting the largest degree nodes,analyze the importance of hubs in maintaining social network links.It is suggested that the network should be more robust when focusing on 10% of the highest degree nodes.Research helps to clarify the self-organizing nature of online social networks.