Research on the Method of Guiding Network Public Opinion on Public Health Emergencies Based on Hybrid Recommendation Algorithm : Taking Microblog Rumors as an Exmaple

Online public opinion about public health emergencies is highly likely to generate extensive public discussion, and online rumors, as part of negative online public opinion, can pose difficulties for public emergency response. To achieve accurate online public opinion guidance, this paper proposes a hybrid recommendation algorithm based on rumor sensitive user profiles and microblogging text profiles. The proposed method first constructs rumor sensitive user profiles and microblogging text profiles with multi-attribute features, and then uses K-means clustering to obtain similar user groups for the constructed rumor sensitive user profiles, effectively reducing the number of microblogging texts in the recommendation candidate set. Finally, by the similarity of the microblogs, we can recommend the correct microblogs of similar people to the disinformation believers, thereby implicitly influencing the misinformation opinions of the disinformation believers and achieving the accuracy and personalization of online opinion guidance. In this paper, we timely capture the trend of online public opinion on public health emergencies from a microscopic perspective, advance the guidance of online public opinion from post-event guidance to mid-event guidance, and formulate a practical method for guiding online public opinion on public health emergencies.