Rise and fall of the global conversation and shifting sentiments during the COVID-19 pandemic

Abstract Social media (e.g., Twitter) has been an extremely popular tool for public health surveillance. The novel coronavirus disease 2019 (COVID-19) is the first pandemic experienced by a world connected through internet. We analyzed 105+ million tweets collected between March 1 and May 15, 2020, Weibo messages compiled January 20 covering six languages (English, Spanish, Arabic, French, Italian, Chinese) represented estimated 2.4 billion citizens worldwide. To examine fine-grained emotions during pandemic, we built machine learning classification models based on deep language to identify in social conversations about COVID-19, including positive expressions ( optimistic, thankful , empathetic ), negative pessimistic, anxious, sad, annoyed denial complicated expression, joking which not explored before. Our analysis indicates rapid increase slow decline volume of regarding all languages. upsurge was triggered combination economic collapse confinement measures across regions belonged except Chinese, where only latter drove conversations. Tweets conveyed remarkably similar emotional states as epidemic elevated status, feelings dominated mixture with anxious / pessimistic conversation surged shifted general optimistic strongest being expressed Arabic tweets, came under control.