Facebook News Feed personalization filter: a case study during the Brazilian elections

The 2018 Brazilian presidential elections were marked by a massive use of social media. Promising to improve user's quality of experience, social media recommend, filter and reorder the posts that will be shown to users. These social media filtering and personalization algorithms determine the flow of posts on the network and shape users' information diets. In this paper, we report measurements and findings from Facebook in the 2018 Brazilian presidential elections. To this aim, a reproducible methodology encompassing a method to identify publishers and candidates alignment and measurements to evaluate the posts selection carried out by Facebook filtering algorithms. The obtained results leverage the understanding of social media influence on the Brazilian elections. This knowledge can be useful in future elections around the world to ensure that the population exercises their right to choose their representatives in a democratic way.