Acessando informação de esporte através do SofaScore

• Ano de defesa: 2022
• Autor(a) principal: Vitor Faria de Carvalho Oliveira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ICX - DEPARTAMENTO DE ESTATÍSTICA; Programa de Pós-Graduação em Estatística; UFMG
• Palavras-chave em Português: banco de dados; coleta; automática; R; SofaScore; web scraping; Estatística – Teses; Ciência de dados – Teses; Banco de Dados – Esportes - Teses; Mineração de dados (Sistemas de recuperação da informação) – Teses
• Link de acesso: http://hdl.handle.net/1843/77330; https://orcid.org/0000-0002-7894-6395

Sports games have become popular, and football being the main one of these sports, followed by around four billion people worldwide. Such popularity has not only provided important cultural transformations, but has also had a large economic impact. Because of the increase in discussions and interest in analyzing soccer matches, it became necessary to provide structured database to universalize access to this data. Along with this need, there is the difficulty of obtaining updated and reliable data for the general population, so that the information can be easily analyzed and studied. Thinking about the previous comments, the present work intends to deliver football data in an easy and practical way to all people, reaching different types of users. The SofaScore site was selected for these reasons and for being an authentic and expert data source for extracting the information. This site is one of the biggest references in consultations on sports matches, having an audience of 22 million people and 22 years of market, for these reasons it was chosen for the extraction. Through the data science tools were selected data from the Brazilian championship of the A series, for being a championship with very active spectators, many rounds and one of the most relevant tournaments in the world. In this work, we use the R software combined with the data collection technique called web scraping, responsible for automatically finding pages, selecting, and extracting the desired content.In this way, a package with several features was developed that helps users extract information of interest in a direct and simple way.