通过对在线社交网络信息的传播过程和传播路径图的介绍,根据研究对象的不同,将学者们近几年提出的信息传播模型分为基于传播路径图节点和基于传播路径图规模两类模型.根据传播路径图节点模型所研究的影响因素的不同将其划分为线性阈值模型及其改进模型、独立级联模型及其改进模型、多因素模型、博弈论模型共四类模型.根据传播路径图规模模型中采用的理论方法的不同将其划分为SIR模型及其改进模型、马尔科夫模型、场强模型、神经网络模型、回归模型共五类模型.文中对九类模型的建立思路、关键点、改进方向等方面进行了详细分析,并根据提出的模型评价指标,对九类模型进行了定量比较.
By the introduction of information propagation process and propagation path graph, this paper divides the information propagation models proposed by scholars in recent years into two categories based on the difference of research objects. One is models based on the nodes of the propagation path graph, and the other is models based on the scale of the propagation path graph. According to the difference of the considered influencing factors, the models based on the nodes of the propagation path graph are divided into four categories : Linear Threshold Model and its improved model, Independent Cascade Model and its improved model, Multi-Factor Model, and Game Theory Model. In view of different theoretical methods, models based on the scale of the propagation path graph are divided into SIR Model and its improved model, Markov Model, Field Strength Model, Neural Network Model, Regression Model. After detailed analysis of the nine models from constructions ideas, key points and future improved direction, this paper then compares the nine models quantitatively based on the proposed evaluationmetrics.