中文摘要：

基于两组实证数据，研究了在线社会网络中用户的优先连接行为，发现虽然对不同的网络优先性指数存在较大的差异，但用户均表现出统计上显著的优先连接行为．基于此构建了一网铬增长模型，考虑了三种场景下的优先性，模型解析解、数值模拟与实证数据三者之间取得了较好的一致．对实诲网络数据进行了统计分析，讨论了两网络度分布统计特性的异同．分析了结构网络之上的用户交流网络，发现用户之耐的平均交流强度在其度值超过邓巴数后迅速减小．

英文摘要：

Based on two empirical datasets this paper studied the preferential linking behavior of users in online social networks and found,that preferential linking is statistically significant for users although there exists distinct difference for the preference exponents of different networks. Based on preferential linking the paper proposed a network growth model incorporating the preference of three aspects. Analytical solution of the model, numerical simulation and empirical data agree well with each other. The paper studied the statistical characteristics of empirical network data and discussed the similarities and differences between the degree distributions of two networks. The paper analyzed the communication network among users on the structure network and found that the mean communication intensity between users rapidly reduces when the degrees of the users exceed the Dunbar＇s number.