Tuning In: a comprehensive analysis of music recommender systems, playlists, and algorithmic fairness

Brussels, 14/12/2023 – A team of researchers, led by Heritiana Ranaivoson (VUB), has just released a report, titled “Tuning In: A Comprehensive Analysis of Music Recommender Systems, Playlists, and Algorithmic Fairness.” This publication, produced as part of the Fair MusE project, delves into the intricate dynamics of music streaming services (MSS), shedding light on their influence on music consumption.

In the last decade, the recorded music industry has witnessed unparalleled growth, largely attributed to the flourishing streaming segment. The researchers highlight the pivotal role of online platforms, especially MSS, in gatekeeping and curating music recommendations.

The report focuses on algorithmic curation, a key facet of the digitization of music through MSS. The authors categorize recommender systems into four types: collaborative filtering, content-based filtering, context-based systems, and hybrid recommender systems. The attention economy paradigm and the growing competition for listeners’ attention in the MSS landscape are thoroughly explored, emphasizing the MSS’ role in the discoverability of music.

The heart of MSS lies in personalized recommendation features, predominantly manifested through curated playlists. The report distinguishes between proprietary playlists, consisting of algorithmic and editorial playlists, and user-generated playlists. It reveals the intricacies of the “algo-torial” logic employed by proprietary playlists, marking a paradigm shift in the way music is discovered and consumed.

One of the most pressing issues addressed in the report is fairness in algorithmic systems. As algorithms significantly shape our daily lives, the tension between fairness and performance becomes increasingly prominent. The authors delve into distinctions such as group fairness vs. individual fairness and awareness-based fairness vs. rationality-based fairness, challenging the existing literature’s reductive approach.

The Fair MusE report aims to provide an interdisciplinary and comprehensive understanding of how platforms and their algorithmic systems function. In a departure from current literature, the authors argue for a broader perspective, asserting that a fairer music ecosystem requires a deeper understanding of biases and strategies embedded in algorithms. By adopting a multi-stakeholder approach, they hope to contribute to a more equitable and transparent music landscape.

Download the report here.

Authors: Heritiana Ranaivoson, Adelaida Afilipoaie, Valdy Wiratama, Dongxiao Li, Saulo Arias Hernández, Jannick Kirk Sørensen, Antoine Henry.

Cover image by BandLab Music on Unsplash