BovDB: A data set of stock quotes for Machine Learning on all companies from B3 between 1995 and 2020

Stock markets are responsible for the movement of huge amounts of financial resources around the world. This market generates a high volume of transaction data, which after being analyzed are very useful for many applications. In this paper we present BovDB, a data set that was built considering as source the Brazilian Stock Exchange (B3) with information related to the years between 1995 and 2020. We have approached the events’ impact on the stocks by applying a cumulative factor to correct prices. The results were compared with public data from InfoMoney and BR Investing, showing that our methods are valid and in accordance with the market standards. BovDB data set can be used as a benchmark for different applications and is publicly available for any researcher on GitHub.