摘要
针对现有社交网络研究未能充分考虑网络拓扑结构的现状,通过提取更多的影响转发行为的特征因素,利用逻辑斯蒂回归模型得到用户间的信息转发概率,进而建立完整的网络拓扑模型。在转发预测模型的基础上,给出网络节点的影响力评估指标,并对PageRank算法进行改进。最后通过在数据集上的试验,试验结果证明给出的拓扑模型能较好的模拟信息在网络中传播过程,改进后的PageRank算法能较好的对节点影响力进行评估和排序。
Since the existing studies don＇t take full consideration of the network topology of social network. In this paper,by extracting more characteristics which will influencethe information transfer behavior,then using logistic regression model to get the probability between two users, thereby establishinga complete network topology model.Base on the information transfer model,then given the influence of the assessment indicators of network nodes, and improve the Page Rank algorithm. The test results on the dataset proved that topology model simulate the process of information disseminationin the network very well,and the improved Page Rank algorithm can better assess and sort the influence of the nodes.
出处
《电子设计工程》
2016年第23期58-61,共4页
Electronic Design Engineering
