论文部分内容阅读
舆论建模跳出传统的基于最近邻(“person-person”)的交互范式,引入次近邻(“personperson-person”)的影响,刻画网络中邻居的邻居对观点改变的作用,提出舆论演化的社会影响级联模型,分析其在可变聚类系数网络上舆论的演化性质.通过调节网络聚类系数,使用异步更新的方式,观察网络集聚特性对舆论演化的影响.结果表明,1)相比于传统的最近邻影响模型,社会影响级联模型的社会强化作用更大,系统更容易达成共识,初始状态中主流观点的影响将被放大;2)舆论演化结果与网络集聚性和初始状态相关:当系统初始状态p_+≠p_,系统观点演化达到稳态后,网络聚类系数越大,越容易产生主流观点;当初始观点p_+=p_时,即正、负力量势均力敌时,系统共识则难以确定.这种情况和现实社会舆论的演化结果符合.
Public opinion modeling jumps out of the traditional interactive paradigm based on “person-person ”, introduces the influence of “personperson-person ”, and describes the effect of neighbors’ neighbors on the change of perspective in the network. The evolutionary social influence cascade model of public opinion evolution and the evolutionary nature of the public opinion on the variable clustering coefficient network are analyzed.The influence of the network clustering characteristics on the evolution of public opinion is observed by adjusting the clustering coefficient of the network and using the asynchronous update mode.The results show that, 1) Compared with the traditional nearest neighbor impact model, the social influence cascade model has a stronger social reinforcing effect and the system is easier to reach a consensus. The impact of the mainstream viewpoints in the initial state will be magnified. 2) The result of the evolution of media and the network agglomeration And the initial state: when the system initial state p_ + ≠ p_, the system viewpoint evolves to steady state, the larger the network clustering coefficient is, the easier it will lead to the mainstream viewpoint; when the initial point p_ + = p_, the positive and negative power When the average is evenly matched, the system consensus is difficult to determine, which is in line with the evolution of realistic public opinion.