Dynamic changes in the spatial-temporal patterns of urban green space topics on Twitter during the COVID-19 pandemic

This study aims to detect spatial-temporal changes in urban green space (UGS) topics pre-, during, and after the peak of the COVID-19 pandemic. Twitter data was selected as data source. Structural topic modelling (STM) was used to identify UGS topics and detect the trends of all topics over time. The inverse distance weighted (IDW) interpolation method was used to show the spatial distributions of all topics over all periods. The research found that the topic Nature observation was the most popular among all topics and showed an increasing trend in topic proportions and dynamic changes in spatial-temporal patterns.