Algorithmic Auditing for Music Discoverability
Автор: MusicTechnologyGroup
Загружено: 2025-11-07
Просмотров: 25
Algorithmic Auditing for Music Discoverability: Engaging Users to Broaden Cultural Diversity in Recommender Systems
Seminar by Lorenzo Porcaro (Sapienza University)
MTG-UPF, November 6, 2025
Abstract:
In this talk Lorenzo presents AA4MD (Algorithmic Auditing for Music Discoverability), a Marie Skłodowska-Curie Actions initiative aiming to explore and mitigate problematic behaviors in music recommendation systems that limit exposure to culturally diverse content. Indeed, as streaming platforms and algorithmic recommenders increasingly mediate how audiences discover music, concerns have emerged around fairness, inclusion, non-discrimination, and transparency. A4MD adopts a human-centred auditing approach: combining qualitative and quantitative methods to understand user experiences; developing a web-based tool for large-scale audits of real recommender systems; and deriving policy recommendations to support more inclusive music discoverability. By involving end users directly in the auditing process, the project seeks not only to identify bias or hidden filters in existing systems but also to pave the way for system designs and policies that amplify under-represented music.
Доступные форматы для скачивания:
Скачать видео mp4
-
Информация по загрузке:
