DESCOBERTA DE CONHECIMENTO SOBRE O SUICÍDIO NO BRASIL EM UMA BASE DE DADOS EXTRAÍDA DO KAGGLE

Abstract

Knowledge Discovery in Databases (KDD) is an increasingly common practice for the analysis of large volumes of data. For this it is necessary to use tools that assist in the search for patterns among the data. Currently it is possible to access large databases either compiled by researchers or made available in government systems. Kaggle is one of these platforms that makes available, among other resources, a huge amount of datasets from several areas. On this platform it was possible to obtain the database with compiled records on suicide in Brazil between the years 2010 and 2019, obtained from DATASUS, a system linked to the Brazilian Ministry of Health. The data analysis was done using the Pandas library of the Python programming language. The qualitative results show that the use of high-performance data analysis libraries is essential to abstract the complexity of the analysis models and ensure the quality of the information generated, allowing the specialist to dedicate himself to the object of study. Data shows that suicide cases in Brazil have been increasing every year, in 2010 the rate of deaths per 100,000 inhabitants was 5.0 in 2019 it jumped to 6.4. The South region has the highest rate of deaths by suicide, in 2019 it was 10.6 per 100,000 inhabitants. The analysis of the results also shows a strong relationship between the underlying cause of suicide with the variables age and state.