群体极化现象存在于大多数在线社交网络信息的传播过程中,对社会观点的形成具有十分重要的影响。作为影响因素之一,社交网络拓扑结构对群体极化现象的影响研究尤为重要。本文基于经典的BA无标度网络模型和群体极化产生机理,提出了改进的局域世界演化模型和群体极化模型。分别基于话题网络和真实网络,探究不同平台网络结构对信息的传播与群体观点形成影响的不同,为社会化网络环境中商务模式的创新提供了理论依据。 Group polarization phenomenon exists in most of the online social network information dissemination process, the formation of social opinion has a very important influence. As one of the influencing factors, the research on the influence of social network topology on group polarization is particularly important. Based on the classical BA scale-free network model and the mechanism of group polarization generation, this paper proposes an improved local world evolution model and population polarization model. Based on the topic network and the real network respectively, this paper explores the different influence of the network structure of different platforms on the information dissemination and group viewpoint formation, which provides a theoretical basis for the innovation of business model in the social network environment.