基于全局和局部结构的微博意见领袖识别研究

Research on Identifying Microblog’s Opinion Leader Based on Global and Local Structure

[目的/意义]准确识别微博意见领袖是成功引导话题舆论导向的关键之一。[方法/过程]在构建用户关系网络的基础上,从节点的信息交互能力和节点的中心性出发,提出了基于全局和局部结构的意见领袖发现模型。以新浪微博中“你愿和导盲犬同行吗”话题为例,进行了实验评估。[结果/结论]基于全局和局部结构的意见领袖识别方法能够体现意见领袖在全局和局部影响力的差异,从而为不同话题下的舆情监控和引导提供更有针对性的建议。

[Purpose/significance]One of the keys to successfully guiding or leading public opinions is to accurately identify“microblog opinion leaders”.[Method/process]Based on the construction of the relationship network between users,starting from the perspective of information interaction ability and centrality of nodes,the paper proposes a new opinion leader discovery model based on global and local structure.Also,the paper conducts an experimental evaluation by taking the topic of“would you like to be with the guide dog”in Sina Weibo as an example.[Result/conclusion]The opinion leader identification method based on global and local structure can reflect the difference of opinion leaders influence in the global and local,so as to provide more targeted suggestions for public opinion monitoring and guidance under different topics.