Abstract:
Over the past two decades, much scholarship has theorized about how highly personalized news media might change the public sphere. But even as algorithmic content filtering has become widespread, social science research has lagged in understanding how such systems work, and how they have altered competitive dynamics between media outlets. Drawing on recent research into recommender systems, this paper examines the Netflix prize, as well as collaborative filtering algorithms deployed by Google and Yahoo. Content recommendation systems strongly advantage the very largest websites over small news outlets, with profound implications for the online news landscape.
