Urban-regional disparities in mental health signals in Australia during the COVID-19 pandemic: a study via Twitter data and machine learning models

Abstract

Abstract This study establishes a novel empirical framework using machine learning techniques to measure the urban-regional disparity of the public’s mental health signals in Australia during the pandemic, and to examine the interrelationships amongst mental health, demographic and socioeconomic profiles of neighbourhoods, health risks and healthcare access. Our results show that the public’s mental health signals in capital cities were better than those in regional areas. The negative mental health signals in capital cities are associated with a lower level of income, more crowded living space, a lower level of healthcare availability and more difficulties in healthcare access.