Objective  To analyze personal and official media micro-blogging on the epidemic of influenza A (H3N2) in 2017 in Hong Kong and to explore problems in risk information dissemination via official micro-blog and to put forward relevant suggestions.

  Methods  Five W (who, says what, in which channel, to whom, and with what effect) methodology, rapid optical screening tool (ROST) mining system, PKUVIS Weibo Visual Analysis System of Peking University and natural language processing and information retrieval (NLPIR) semantic analysis system were used in the study. We adopted content analysis to classify the information and content words which being put out by different micro-bloggers. Then we carried out a sentiment analysis and forwarding path analysis.

  Results  We identified 130 micro-blogs on influenza A (H3N2) epidemic in Hong Kong during the period from July to September 2017, of which, 43.85% and 41.54% were personal and official media micro-blogs and only 7 of which were micro-blogs of governmental agencies. Emotional information of the micro-blogs was time-concentrated and negative. The contents of official media micro-blogs released by People's Daily on the 5th and 6th of August 2017 showed a typical hierarchical structure, including small-scale second-time forwarding and the third-time forwarding, and comparison analyses on comments revealed that more than 90% of the public were fully convinced about the micro-blogs' information but the rate decreased later on. Among the public, 8.09% and 6.44% reported changes in behavioral intention after reading the information and about 5% reported fear and panic emotion. Sentiment content analysis identified higher proportions of happy, nice, fear, and evil emotions.

  Conclusion  The official media should maintain the authority, precision, and objectivity of their micro-blogs and use words cautiously and avoid exaggerated news headline. The results also suggest that information literacy of the public needs to be improved and government agencies should strengthen monitoring and guiding on micro-blogs about public health emergency.