[目的/意义]微博舆情的形成是一个多层面、错综复杂的过程,可采用超网络描述和揭示。[方法/过程]构建了包含用户-观点-情感-时序阶段4层子网的超网络模型,并将该模型应用于“毒疫苗”事件这一具体实例中。[结果/结论]研究表明,舆情主题发现超网络模型的子网分析可揭示每层子网的特征信息,超边分析可用于舆情预警分析、舆情主题挖掘及舆情主题演化分析。[局限]下一步研究将从细化指标、多重验证两个方面对模型进行完善。
[ Purpose/significance] The public opinion formation in micro-blog is a muhi-faceted and complex process, which can be described and revealed by the super-network. [ Method/process ] This paper establishes a super-network including 4 subnets of users, opinion, sentiment, and timing phase, and applies the model to a specific instance of micro-blog public opinion. [ Resuit/conclusion ] Each sub-network of public opinion topic mining can reveal its feature information, and super-edge analysis can be used in public opinion early warning, public opinion topic mining and public opinion topic evolution. [ Limitations ] The next step is to improve the model from aspects of parameters refinement and multiple validations.