Abstract: Purpose/Significance To provide decision support for accurate monitoring and classification guidance of network public opinion of public health emergencies. Method/ProcessThe web crawler technology was used to obtain the comment data and the characteristic data of comment users on Weibo related to public health emergencies, and the sentiment analysis was carried out to calculate the emotional tendency of the review data. Correlation analysis was used for calculated emotional degree of emotional inclination and user characteristics. Result/Conclusion User comments emotional tendency and six user characteristics related to user comments emotional tendency such as gender, age, number of fans, number of likes, introduction and label were used as content to construct user portraits of Weibo comments public health emergencies.