基于LDA-DLNB网络舆情识别的应急管理智库决策研究OA北大核心CHSSCDCSSCICSTPCD

[Purpose/Significance]Network public opinion information has a significant communication in-fluence in emergency response.Improving public opinion guidance is one of the important functions of new think tanks with Chinese characteristics,and effectively identifying network public opinion is the key for emergency management think tanks to enhance their comprehensive research and judgment abilities.[Method/Process]From the perspective of constructing emergency management think tanks,this paper focuses on identifying potential key hotspots and emotional structures in network public opinion information related to emergencies.It constructs the LDA-LDNB network public opinion information mining model using the machine learning method,and conducts model validation and empirical analysis based on the textual information from network public opinion regarding emergencies.[Result/Conclusion]The results demonstrate that the model constructed in this paper can effectively identify the themes,emotions and evolutionary features of network public opinion information during emergen-cies,provide methods and decision-making support for emergency management think tanks to improve their abil-ity to participate in network public opinion research and judgment for emergency management decision-making services,and offer a new way of thinking for constructing a data-driven emergency management think tank.