Understanding musical success beyond hit songs: characterization and analyses of musical careers

• Autor(a) principal: Danilo Boechat Seufitelli
• Data de Publicação: 2023
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Repositório Institucional da UFMG
• Texto Completo: http://hdl.handle.net/1843/61054; https://orcid.org/0000-0002-0155-7631
• topic: Musical success; Hot streaks; Hit song science; Artist careers; Computação – Teses; Computação – Cultura brasileira - Teses; Computação – Música – Teses; Sucesso profissional – Cantores brasileiros - Teses; Indústria musical - Brasil – Teses.

Musical careers are dynamic, expressive, and fundamental to an artistic and cultural experience. Its dynamism can be seen through the many changes in recent decades concerning music consumption: we have moved from vinyl, cassettes, and CDs to the streaming platforms that are here to stay. Streaming brought with it the high availability of data associated with music consumption and listener preference. With such data, we can extract relevant knowledge, such as what can lead some songs to success and others not. In this scenario, a critical study area called Hit Song Science emerged, whose main objective is to reveal the music industry’s success dynamics. Collecting hit songs can lead artists to experience periods of success far beyond the ”ordinary” periods known as Hot Streaks. In this sense, understanding how the different profiles of artists stand out and reach their most successful periods can be crucial for the music industry, which deals with the constant natural evolution of the market and needs to reinvent itself to satisfy the desires of its consumers: connect successful music and artists. Hence, our objective in this thesis is to identify the characteristics that lead artists to reach their most successful periods. We first conducted an extensive literature review to identify the main definitions of success, characteristics, and algorithms used. As a result, we propose a taxonomy and a generic flow for Hit Songs Science. Next, we study how music consumption evolved in the Brazilian market, analyzing the transition period from the physical to the digital era. We found that artists’ most successful periods tend to cluster in time, and we identified periods of Hot Streaks. Furthermore, we detected that some musical genres have significant patterns for both eras. We also performed a profile analysis that revealed three different groups in both eras: Spike Hit Artists (SHA), Big Hit Artists (BHA), and Top Hit Artists (THA), which acted as class descriptors of successful artists. Finally, we discovered that part of the Brazilian population with access to music streams prefers to consume music by Brazilian artists, regardless of the era. Finally, we investigate a possible regularity in exploring different topics in artistic careers before they experience their first period of above-normal success. For this, we propose a data-based methodology to analyze how artists spread their interests (Exploration) and focus their attention (Exploitation) on different musical topics (e.g., musical genres) while achieving peaks of success (Hot Streaks). Hence, we measure the entropy of artists’ careers, which results in the degrees of Exploration and Exploitation. The Exploration phase indicates that artists diversify their work topics. In contrast, in the Exploitation phase, there is a specific definition of the work focuses, refining its capabilities over time. We identify the musical topics by detecting the community in a complex modeling network of artists, songs, and their genres. Then, we measure the entropy degree (in terms of Exploration and Exploitation), quantifying it in three periods: before, during, and after your Hot Streaks. Results show that artists explore more topics before hitting their first wave of success and during their most successful period. Afterward, they narrow the range of topics, delving deeper into just a few. Such findings are relevant to identifying and nurturing creative talent in the music industry.