Simulating the Algorithm Store: Multistakeholder Impacts of Recommender Choice
Anas Buhayh, Elizabeth McKinnie, Clement Canel, Robin Burke · 2025
Recommender systems play an essential role in connecting users with items.Traditionally, research in this field has focused on refining recommendation algorithms within monolithic systems that reside in a single platform.We are exploring alternative architectures in which users have a choice over recommendation algorithms.In this work, we use simulation grounded in real-world data to explore the impact of such alternative designs on recommendation stakeholders.We show that consumers of niche items and producers of such items can both benefit from algorithmic choice.