论文部分内容阅读
舆论领袖对网络舆情的发生和发展有重要的引导作用,可以影响网络舆论的走向,识别其中的舆论领袖具有重要的现实意义。首先建立了集社交、环境、心理和观点四层子网描述的网络舆论超网络模型,将现有的对网络舆论进行的单层社会网络分析扩展为超网络研究;在超网络模型基础上,提出了一种新的超边排序算法(SuperEdgeRank),通过该算法对社交子网中各个舆论主体参与形成的超边进行排序,进而挖掘出网络舆论领袖。在该算法中,分别对环境子网中信息传播影响度、心理子网中不同心理动机之间转化关联度和观点子网中不同观点的相似度等参数进行了计算。最后通过实例分析证明了该方法的可行性,具有很强的理论指导意义和实际应用价值。
Public opinion leaders play an important guiding role in the occurrence and development of online public opinion, which can influence the trend of online public opinion and identify the public opinion leaders therein. First of all, we set up a super-network model of network public opinion which is described by the four sub-networks of social, environmental, psychological and opinion. The existing single-layer social network analysis of public opinion of the network is extended to super-network research. Based on the super-network model, A new super-edge ranking algorithm (SuperEdgeRank) is proposed. Through this algorithm, the super-edges formed by the participation of various public opinion entities in the social sub-networks are sorted, and then the network opinion leaders are discovered. In the algorithm, parameters such as the degree of influence on information dissemination in environmental subnets, the transformation relevance between different psychological motivations in mental subnets, and the similarity of different views in subjective views were calculated. At last, the feasibility of this method is proved by the example analysis, which has strong theoretical guidance and practical value.