Analysis of sentiments conveyed through Twitter concerning COVID-19

Due to the social and economic fallout from the COVID-19 pandemic, we sought to gauge the attitudes of social network users, in this case, Twitter, towards the topic using a sentiment analysis approach. We collected 178,683 tweets using the Twitter API based on queries for the high-frequency hashtag #covid19. After the preprocessing step, we classified them in a binary way (positive and negative) and according to their intensity (valence) using the VADER model and then the NRCLex dictionary, which allows us to classify feelings according to their affective class. The results suggest that overall, the feelings detected through the tweets are positive. In addition, users seem to be interestedin the pandemic as a trend rather than as a topic related to other social or economic aspects.