Author:
Santos Rafael V. M.,Comarela Giovanni V.
Abstract
As a highly data-driven application, recommender systems can be affected by data distortions, culminating in unfair results for different groups of data, which can be a reason to affect system performance. Therefore, it is important to identify and resolve issues of unfairness in referral scenarios. We therefore developed an equity algorithm aimed at reducing group injustice in recommender systems. The algorithm was tested on two existing datasets (MovieLens and Songs) with two user clustering strategies. We were able to reduce group unfairness in both data sets by considering the two clustering strategies.
Publisher
Sociedade Brasileira de Computação
Reference34 articles.
1. Barocas, S. and Selbst, A. D. (2016a). Big data’s disparate impact. Calif. L. Rev.. California Law Review, 104(IR):671.
2. Barocas, S. and Selbst, A. D. (2016b). Big data’s disparate impact. California Law Review, 104(3):671–732.
3. Berk, R., Heidari, H., Jabbari, S., Joseph, M., Kearns, M. J., Morgenstern, J., Neel, S., and Roth, A. (2017). A convex framework for fair regression. CoRR, abs/1706.02409.
4. Beutel, A., Chi, E. H., Cheng, Z., Pham, H., and Anderson, J. (2017). Beyond globally optimal: Focused learning for improved recommendations. In Proceedings of the 26th International Conference on World Wide Web, WWW 2017, Perth, Australia, April 3-7, 2017.
5. Biega, A. J., Gummadi, K. P., and Weikum, G. (2018). Equity of attention: Amortizing individual fairness in rankings. CoRR, abs/1805.01788.