Aplicação de análise de sentimentos para percepção da saúde mental dos usuários do Twitter no início da pandemia de covid-19

Abstract

Every day, the digital sphere is flooded with an immeasurable variety of texts - news, scientific articles, comments, reviews, among others. These written narratives hold the power to express opinions and emotions, with social networks serving as a significant stage for such expression. However, due to the immense textual production every instant, many of these pieces of information end up not being properly analyzed. This study aims to implement a sentiment analysis algorithm to probe publications made on Twitter during the COVID-19 pandemic, focusing especially on the keywords: Brazil and Coronavirus.Mental health has been emerging as a pressing concern during the COVID-19 pandemic, and social networks have played a paramount role in this scenario. However, it is crucial to recognize that this correlation is intricate, possibly leading to both positive and negative impacts on individuals’ mental health. The sentiment analysis conducted in this study offered a deeper understanding of the pandemic’s impact on Twitter users. By analyzing thousands of tweets, it became possible to capture the collective sentiment during the crisis and identify trends, patterns, and oscillations of feelings over time. This evaluation allowed us to translate events and experiences from the physical world to the Twitter environment, unveiling the magnitude of the psychological impact of the pandemic. It was possible to discern the presence of feelings such as fear, anxiety, sadness expressed by the platform’s users.