|
Abstract:
|
Music is often a powerful means to bring societies together during a crisis, from protest marches invoking familiar hymns to quarantined citizens singing together from their balconies. Is it possible, then, to use a city’s music preferences to learn about the way its residents cope with and potentially recover from such a crisis, and if so, how?
This case study uses music streaming data from Spotify and anomaly detection methods to explore these behavioral patterns. By analyzing anomalies in a city’s streaming behavior, we aim to identify major events occurring in that city and the effects of those events on streaming behavior. We focus on cities particularly impacted by the resurgence of the Black Lives Matter movement in the summer of 2020, noting the behavior of streaming data can become noticeably different after a flashpoint event, such as the death of George Floyd for the city of Minneapolis. We show how similar effects manifest throughout the United States using standard anomaly detection algorithms. Ultimately, we correlate anomalies in a city’s music streaming behavior with events occurring in that city to generate insights on music as a means of resilience in society.
|
