Investigating the civic emotion dynamics during the COVID-19 lockdown: Evidence from social media

Shanghai implemented strict "citywide static management" under the "zero-COVID" policy when the pandemic re-emerged in 2022, sparking a unique public emotion wave. This research examines the correlations between the spatial-temporal heterogeneity of the negative emotion and the micro-level built environment based on civic emotion extracted from social media from the spatial justice perspective. Specifically, it obtains Weibo posts during March, April and May 2022 to cover the pre- and during lockdown periods to identifies negative emotions from public voices. Secondly, it expands commonly used negative emotions with words related to a unique emotion – helpless during Shanghai's lockdown. It then applies natural language processing (NLP) to grade the negative emotion across Shanghai. Third, it describes built environment, housing economy, socio-demographic and pandemic spread data as independent variables. Lastly, utilizing Ordinary Least Squares (OLS), Spatial Lag and Geographically Weighted Regression (GWR) models, it explores the correlations between emotion and the independent variables, to analyze the influential factors of negative emotions as expressed on social media. Our findings reveal significant spatial heterogeneity and temporal dynamics of the correlation between local emotions and complex socio-economic attributes, critically supporting the urgency and importance of spatial justice. This data-driven approach retraces the local voices and discloses negative emotions under lockdown measures. The study contributes to the understanding of urban resilience and emotional well-being in the face of public health crises, offering insights into sustainable urban planning and policy-making for resilient cities.