2021-09-21 16:47:55 - Atualizado em 2021-09-21 16:48:47

Ciência de Dados para Eleições: Analyzing spatio-temporal voting patterns in Brazilian elections

Por Ricardo Marcacini

This paper proposes a simple data science pipeline to analyze Brazilian elections at the municipal level, in the spatial and temporal dimensions. The pipeline consists of five steps: (i) data selection; (ii) data preprocessing; (iii) identification of spatial patterns, where we seek to understand the role of space in the election results employing spatial autocorrelation techniques; (iv) identification of temporal patterns, in which we explore similar trends of votes over the years applying a clustering method; and (v) evaluation of results. We study the presidential elections from 1998 to 2018 focusing on the most relevant left and right parties for the period: Workers' Party (PT) and the Brazilian Social Democracy Party (PSDB). We also analyze the congressman election data concerning parties ideologically to the left and right in the political spectrum. From the obtained results, we found the existence of spatial dependence in every electoral year investigated. Furthermore, despite the changes in the political-economic context over the years, neighboring cities presented similar voting behavior trends.


Venue: Journal of Data Information and Management 

Autores: Lucas Mantovani, Tiago Pinho da Silva, Antonio R. S. Parmezan, Gustavo E. A. P. A. Batista

DOI: 10.5753/jidm.2021.1932