Pandemia na era da (des)informação: a ciência de dados na análise de informações falsas no contexto da Covid-19 no Brasil

Abstract

This work uses an exploratory study proposing to identify the misinformation about COVID-19 with data mining techniques to look for patterns in messages published on the virtual social network - Twitter with reference to the keywords: Vaccination Campaign, Pandemic, COVID-19. Analysis of posts made on Twitter in order to classify Tweets if misinformation is present. Investigating the use of analysis in verification agencies to verify information. Development of the database with the tweets, extracted by periods. Use of existing tools to verify information, user profile, data format initially in Comma-Separated Values (CSV) generated with the Python programming language for the data processing steps in the API. Therefore, initially a bibliographic review was carried out that sought to contextualize the social network used in Brazil. The publications refer to the month of October to November 2021, and to the month of March and April 2022, made in Portuguese, whose geographic coverage was the Brazilian territory.