COVID-19疫情防控中的中国公众舆情时空演变特征

COVID-19疫情是全球面临的重大公共卫生危机。客观认识疫情期间的公众舆情响应和区域差异,对于提高重大公共卫生事件的政策调控和科学治理具有现实意义。本文以新浪微博为数据源,基于潜在狄利克雷分配主题模型和随机森林算法构建了主题抽取和分类模型,识别微博文本中的13个舆情话题,并从数量、空间、时间、内容等方面分析了2020年1月9日—3月10日在湖北省、京津冀、长三角、珠三角、成渝等城市群及沿边口岸等重点区域分布特点。结果表明：中国公众的响应总体是理性和积极的,但各舆情话题在区域内部的空间分布差异明显。各区域热点分布中,京津冀以首都北京为中心,长三角以上海为中心,辅以南京、杭州等热点,珠三角以广州、深圳为两核,湖北省以武汉为中心。建议应持续加强重点区域的疫情舆情关注和因地制宜的差异化精准响应。

Spatio-temporal evolution and regional differences of the public opinion on the prevention and control of COVID-19 epidemic in China

The COVID-19 epidemic is a global public health crisis. It is of practical significance to objectively understand the public's responses and regional differences in order to improve policy control and scientific governance during major public health threats. In this study, a topic extraction and classification model was constructed based on the Latent Dirichlet Allocation topic model and the Random Forest algorithm. Thirteen topics were identified about public opinion in the Chinese SINA microblog from January 9 to March 10, 2020. The regional distribution characteristics were explored in terms of the amount, space, time sequence, and content in major urban agglomerations including Hubei Province, Beijing-Tianjin-Hebei urban agglomeration, Yangtze River Delta, Pearl River Delta, Chengdu-Chongqing region, and some border ports of China. The results showed that the spatio-temporal distribution of public opinion is related to the severity of the epidemic, degree of population aggregation, and level of economic development. The response of Chinese people is rational and positive, and the spatial distribution within these regions is obviously different. Among the regional hotspots, Beijing-Tianjin-Hebei region is centred on Beijing; the Yangtze River Delta is centred on Shanghai, followed by Nanjing, Hangzhou and other hotspots; the Pearl River Delta is centred on Guangzhou and Shenzhen; and Hubei Province is centred on Wuhan. The time series of topics in each region are synchronously related, but there are differences in timing sequence and periodic fluctuation in response time and intensity. The imbalance of resource allocation caused by the sharp rise of relief information in the short term is prominent, and the differences in response policies of various urban agglomerations combined with regional characteristics are not obvious. We should continue to focus on public opinion on epidemic situations in key areas and accurately respond to local regions according to its actual conditions.