Análise de padrões temáticos e emocionais em letras de músicas periféricas brasileiras utilizando ciência de dados

Abstract

This dissertation investigates the use of Data Science techniques for analyzing from Brazilian peripheral genres lyrics. As a form of cultural expression, music has a unique ability to capture and convey the complex emotions and realities of a people. In this context, this dissertation analyzes song lyrics from Funk, Rap and Samba through Data Science to identify emotion and thematic patterns present in such compositions. The technical contribution is also expanded with two new datasets collected from online sources and subsequently processed for easy and effective use. Furthermore, this study offers new perspectives on the cultural expression of Brazilian peripheral regions, then showing how technology can be applied to understand and interpret complex textual data. The dissertation concludes that the intersection between music and technology not only enriches the field of Computer Science, but also highlights the potential of computational techniques to broaden horizons for cultural and social investigations.