Is it possible to learn about people without hurting their privacy?

Abstract

This article briefly describes the concept of individual privacy in the Brazilian General Data Protection Law (Lei Geral de Proteção de Dados - LGPD) while discussing personal data collection for profiling creation in modern machine learning systems. The article outlines a few computational techniques for data anonymization and shows that, by associating these techniques with the collection/learning binomial, we can strengthen the publication and study of data while guaranteeing the privacy of data holders.