How Algorithmic Confounding in Recommendation Systems Increases Homogeneity and Decreases Utility
This addresses a critical issue for users and developers of recommendation systems, as it highlights a feedback loop that reduces diversity and effectiveness, making it incremental by building on known concerns about bias in such systems.
The study tackled the problem of algorithmic confounding in recommendation systems, showing through simulations that using confounded data homogenizes user behavior without improving utility.
Recommendation systems are ubiquitous and impact many domains; they have the potential to influence product consumption, individuals' perceptions of the world, and life-altering decisions. These systems are often evaluated or trained with data from users already exposed to algorithmic recommendations; this creates a pernicious feedback loop. Using simulations, we demonstrate how using data confounded in this way homogenizes user behavior without increasing utility.