中文摘要：

创新性地将复杂系统多主体建模方法应用到集社交、环境、心理和观点等子网的舆情超网络模型中,通过对该超网络模型中各子网间不同属性节点（即异质Agent）交互规则设计,建立基于超网络的多主体舆情演化模型,深度刻画网络舆情的演化规律,即模拟人-信息-观点之间微观意见如何交互并涌现出宏观网络舆情的过程;应用舆情超网络拓扑结构相关测度指标,设置多主体舆情演化模型中各属性参数值,如网民活跃度（节点超度）、信息影响范围（超边间距离）和舆论领袖权威性（超边重叠度）等,并重点分析舆情演化过程中政府抢占舆论先机的重要作用,以及舆论领袖对大众网民的引导作用和网民发表观点形成舆论场的控制作用。最后,运用NetLogo仿真平台对具体舆情实例进行仿真,结果与现实舆情演化情况相符,证明了模型的合理性与有效性。

英文摘要：

We apply the multi-agent modeling framework in the Opinion SuperNetwork（OSN）,which includes subnetworks of social,psychological,environmental and perspective elements.We first establish a multiple-agent public opinion evolution model through the design of interactive rules of nodes with different characteristics among all the subnetworks to deeply display the evolution rule of online public opinion.We simulate the processes of how the microscopic views among people,information and keywords interact and form the macroscopic public opinion;We then apply the related measurement indices of supernetwork topology and set the attribute value of some indices of multi-agent model,such as netizen＇s activity（node superdegree）,influential scope of information（superedge-superedge distance）and authority of opinion leader（superedge overlap）.We also analyze the important role of government in seizing the opportunity of public opinion and opinion leaders＇ guiding role to the internet users as well as the control on the internet users＇ issuing views to form public opinion field.Finally,we use NetLogo simulation platform to simulate the specific case of public opinions.The result is consistent with the actual public opinion evolution,demonstrating the rationality and validity of the model.