The purpose of this paper is to present an approach for forecasting mental health conditions and emotions of a given population during the COVID-19 pandemic in Argentina based on social media contents.
Mental health conditions and emotions are captured via markers, which link social media contents with lexicons. First, the authors build time series models that describe the evolution of markers and their correlation with crisis events. Second, the authors use the time series for forecasting markers and identifying high prevalence points for the estimated markers.
The authors evaluated different forecasting strategies that yielded different performance and capabilities. In the best scenario, high prevalence periods of emotions and mental health issues can be satisfactorily predicted with a neural network strategy, even at early stages of a crisis (e.g. a training period of seven days).
This work contributes to a better understanding of how psychological processes related to crises manifest in social media, and this is a valuable asset for the design, implementation and monitoring of health prevention and communication policies.
Although there have been previous efforts to predict mental states of individuals, the analysis of mental health at the collective level has received scarce attention. The authors take a step forward by proposing a forecasting approach for analyzing the mental health of a given population at a larger scale.
