基于邻居节点间相互影响和改进概率的社交网络信息传播模型

摘要

One of current focus of spreading dynamics research analses the propagation of information in the specific network based on the epidemic dynamics model.According to the characteristics of propagation in social network,this paper added a kind of new node named disguising node based on the susceptible-infectious-recovered (SIR) model,and proposed a model named susceptible-disguising-infectious-recovered (SDIR) to describe the propagation better in social network.Considering the mutual influence of neighbor nodes,it defined three propagation probability functions to improve the SDIR model.The results show,by simulating propagation under different conditions,that information cannot cover the whole network,and Twitter performs better than Sina Micro-blog in efficiency of propagating.Also,the initial infection probability has a significant influence in the information propagation.%目前网络传播动力学的研究焦点之一是以经典的传染病动力学模型为基础,研究特定网络的信息传播规律.针对社交网络中信息传播的特点,在传统的SIR模型基础上,通过加入新的一类假免疫节点,建立了新的SDIR模型.考虑到邻居节点间的相互影响,通过定义三个传播概率函数,对SDIR模型作了改进,得到了更加符合社交网络特点的传播模型.对比不同条件下信息传播的过程,实验证明了信息不能覆盖全网络,Twitter比新浪微博有更好的信息传播效率的推测,并发现初始传播概率会对信息传播有重要影响.