Projects per year
Abstract
As recommender systems are prone to various biases, mitigation approaches are needed to ensure that recommendations are fair to various stakeholders. One particular concern in music recommendation is artist gender fairness. Recent work has shown that the gender imbalance in the sector translates to the output of music recommender systems, creating a feedback loop that can reinforce gender biases over time.
In this work, we examine that feedback loop to study whether algorithmic strategies or user behavior are a greater contributor to ongoing improvement (or loss) in fairness as models are repeatedly re-trained on new user feedback data. We simulate user interaction and re-training to investigate the effects of ranking strategies and user choice models on gender fairness metrics. We find re-ranking strategies have a greater effect than user choice models on recommendation fairness over time.
In this work, we examine that feedback loop to study whether algorithmic strategies or user behavior are a greater contributor to ongoing improvement (or loss) in fairness as models are repeatedly re-trained on new user feedback data. We simulate user interaction and re-training to investigate the effects of ranking strategies and user choice models on gender fairness metrics. We find re-ranking strategies have a greater effect than user choice models on recommendation fairness over time.
| Original language | English |
|---|---|
| Pages | 884-889 |
| Number of pages | 6 |
| DOIs | |
| Publication status | Published - 8 Oct 2024 |
| Event | 18th ACM Conference on Recommender Systems - Petruzzelli Theater, Bari, Italy Duration: 14 Oct 2024 → 18 Oct 2024 Conference number: 18 https://recsys.acm.org/recsys24/ |
Conference
| Conference | 18th ACM Conference on Recommender Systems |
|---|---|
| Abbreviated title | RecSys 2024 |
| Country/Territory | Italy |
| City | Bari |
| Period | 14/10/24 → 18/10/24 |
| Internet address |
Bibliographical note
Publisher Copyright:© 2024 Copyright held by the owner/author(s).
Keywords
- recommender systems
- user choice models
- re-ranking
- artists
- music
- gender
- fairness
- bias
Fields of Science and Technology Classification 2012
- 102 Computer Sciences
- 107 Other Natural Sciences
Access to Document
Other files and links
Projects
- 1 Active
-
EXDIGIT: Excellence in Digital Sciences and Interdisciplinary Technologies
Horejs-Höck, J. (Principal Investigator), Hüsing, N. (Principal Investigator), Tscheligi, M. (Co-Investigator), Bauer, C. (Co-Investigator), Demir, C. (Co-Investigator), Dong, Y. (Co-Investigator), Faria Costa, J. (Co-Investigator), Graf, F. (Co-Investigator), Grilnberger, M. (Co-Investigator), Hannibal, G. (Co-Investigator), Katzlberger, D. (Co-Investigator), Leschke, N. C. (Co-Investigator), Linortner, M. (Co-Investigator), Maislinger, L. (Co-Investigator), Mocnik, F.-B. (Co-Investigator), Pallas, F. (Co-Investigator), Unterberger, E. (Co-Investigator), Williams, M. C. (Co-Investigator), Zimmermann, G. (Co-Investigator), Carrozzo, A. E. (Co-Investigator), Mehran, N. (Co-Investigator) & Dabiri, Z. (Co-Investigator)
1/04/22 → 31/03/28
Project: Research
-
Fairness in algorithmic decision-making: The effects of bias mitigation strategies in music recommender systems
Bauer, C. (Invited speaker)
20 Nov 2024Activity: Talk or presentation › Guest lecture › science to science / art to art
-
It’s not you, it’s me: the impact of choice models and ranking strategies on gender imbalance in music recommendation
Ferraro, A. (Presenter), Ekstrand, M. (Presenter) & Bauer, C. (Presenter)
15 Oct 2024Activity: Talk or presentation › Poster presentation › science to science / art to art
File