A Caracterização de Mulheres na Música Brasileira: Identificando Padrões com Processamento de Linguagem Natural

Abstract

This study investigates the presence of gender biases in Brazilian song lyrics since the mid-20th century. The analyses, conducted using NLP techniques, reveal that women are often described in a stereotypical manner, primarily based on their physical appearances. When characterized in terms of personality and emotions, the emphasis is placed on adjectives suggesting emotional instability. Furthermore, in certain musical styles, actions related to women are frequently associated with sexual acts. The findings regarding professions align with social data provided by Brazilian research institutes. The conclusions of this study are pioneering in the musical field but corroborate prior studies on sexism identification conducted in other domains, highlighting the need for critical reflection on the perpetuation of gender discrimination in society.