A Fairness-Aware Multi-Objective Recommendation Framework for Mitigating Creator Popularity Bias and Balancing User Utility and Engagement

Recommendation systems increasingly function as attention-allocation mechanisms, making creator exposure an important algorithmic outcome. Popularity bias can emerge through feedback loops in which historical visibility produces interactions, interactions influence ranking, and ranking produces additional visibility. Optimizing only for immediate engagement can therefore amplify concentration and reduce opportunities for emerging or long-tail creators. This review argues for a multi-objective approach in which relevance and engagement remain essential but are balanced with user utility, diversity, novelty, and creator fairness. The proposed Fairness-Aware Multi-Objective Recommendation (FAMOR) framework integrates popularity-aware scoring, exposure-aware ranking, cold-start exploration, multi-objective optimization, and continuous monitoring. The framework does not assume that all creators should receive identical exposure; instead, it emphasizes fair opportunity subject to relevance, quality, and user context. Future empirical studies should test the framework using realistic exposure logs and online experiments, with particular attention to causal effects and long-term ecosystem outcomes. Such research can help move recommender systems from purely engagement-centered optimization toward responsible personalization that benefits users, creators, and platforms simultaneously.

Abdollahpouri, H., Burke, R., & Mobasher, B. (2017). Controlling popularity bias in learning-to-rank recommendation. Proceedings of the 11th ACM Conference on Recommender Systems. Abdollahpouri, H., Burke, R., & Mobasher, B. (2019). Managing popularity bias in recommender systems with personalized re-ranking. Proceedings of the AAAI Conference on Artificial Intelligence. Burke, R. (2017). Multisided fairness for recommendation. FATREC Workshop on Fairness, Accountability, and Transparency in Recommender Systems. Ekstrand, M. D., Burke, R., & Diaz, F. (2018). Fairness and discrimination in recommendation and retrieval. Proceedings of the 11th ACM International Conference on Web Search and Data Mining. Fabbri, F., Bonchi, F., Boratto, L., & Castillo, C. (2021). Exposure inequality in recommender systems: A review and research agenda. ACM Computing Surveys. Ge, Y., Liu, S., Fu, Z., Tan, J., Li, Z., & others. (2021). Towards long-term fairness in recommendation. Proceedings of the ACM International Conference on Information and Knowledge Management. Hardt, M., Price, E., & Srebro, N. (2016). Equality of opportunity in supervised learning. Advances in Neural Information Processing Systems. Joachims, T., Swaminathan, A., & Schnabel, T. (2017). Unbiased learning-to-rank with biased feedback. Proceedings of the 10th ACM International Conference on Web Search and Data Mining. Jannach, D., Zanker, M., Felfernig, A., & Friedrich, G. (2010). Recommender Systems: An Introduction. Cambridge University Press. Kunaver, M., & Požrl, T. (2017). Diversity in recommender systems: A survey. Knowledge-Based Systems, 123, 154–162.