Beyond Parity: Fairness Objectives for Collaborative Filtering
This work addresses fairness for users in recommender systems, particularly minority groups, but is incremental as it builds on existing fairness objectives.
The paper tackled fairness issues in collaborative-filtering recommender systems by proposing four new fairness metrics to address biases from historical data, and experiments showed these metrics better measure fairness and reduce unfairness compared to baselines.
We study fairness in collaborative-filtering recommender systems, which are sensitive to discrimination that exists in historical data. Biased data can lead collaborative-filtering methods to make unfair predictions for users from minority groups. We identify the insufficiency of existing fairness metrics and propose four new metrics that address different forms of unfairness. These fairness metrics can be optimized by adding fairness terms to the learning objective. Experiments on synthetic and real data show that our new metrics can better measure fairness than the baseline, and that the fairness objectives effectively help reduce unfairness.