Mitigating Demographic Biases in Social Media-Based Recommender Systems

As a growing proportion of our daily human interactions are digitized and subjected to algorithmic decision-making on social media platforms, it has become increasingly important to ensure that these algorithms behave in a fair manner. In this work, we study fairness in collaborative-filtering recommender systems trained on social media data. We empirically demonstrate the prevalence of demographic bias in these systems for a large Facebook dataset, both in terms of encoding harmful stereotypes, and in the impact on consequential decisions such as recommending academic concentrations to the users. We then develop a simple technique to mitigate bias in social media-based recommender systems, and show that this results in fairer behavior with only a minor loss in accuracy.