Year: 2025
Type: article
Abstract: In the competitive streaming era, understanding regional music consumption is vital, particularly in culturally diverse nations like Brazil. While extensive research exists on national music trends, prior studies have overlooked the nuanced local variations in musical tastes and acoustic characteristics. This study addresses that gap with a data-driven methodology examining genre preferences and acoustic attributes across Brazilian regions using Spotify Charts data from 2022–2023. Our computational approach involved constructing bipartite genre-city networks, applying backbone extraction to identify key preferences, and using temporal analysis to assess their stability. We employed clustering techniques to uncover regional acoustic patterns independent of genre and association rules mining to pinpoint common and exceptional listening behaviors. Finally, we analyzed artist similarity to determine the influence of geography versus genre. Our findings reveal that while core genre preferences remain largely stable across regions and time, significant distinctions emerge from acoustic analysis. We identified distinct city clusters with unique sonic profiles, defined by variations in attributes like liveliness, speechiness, and valence. This demonstrates that regional identity is not solely shaped by genre. Furthermore, our analysis shows that artist similarity is strongly influenced by cultural and geographical proximity. Although many listening behaviors are shared nationally, unique patterns appear in specific clusters, especially during major holidays. This study contributes a robust computational framework for modeling music consumption at a granular geographical level. The results provide valuable insights for artists, digital platforms, and industry professionals, highlighting how regional musical identities are shaped by a complex interplay of genre, acoustic features, and cultural factors.
Source: Journal on Interactive Systems
Authors: Filipe A. S. Moura, Carlos H. G. Ferreira, Helen C. S. C. Lima
Institution: Universidade Federal de Ouro Preto
Language: English
Topic: Music and Audio Processing
Subfield: Signal Processing
Field: Computer Science
Domain: Physical Sciences
SDG: Reduced inequalities
Open Access status: diamond
OpenAlex ID: W7084113355