The research presents CP-FairRank, a re-ranking algorithm that integrates fairness constraints from both consumer and producer perspectives. It aims to improve fairness in recommender systems, which are widely used in various applications like online marketing and music discovery. The framework is adaptable and can be applied to different fairness settings based on group segmentation and recommendation model selection. Empirical validation on eight datasets showed that the proposed strategy could enhance both consumer and producer fairness without significantly compromising recommendation quality. The study highlights the role of algorithms in mitigating data biases.
Publication date: 2 Feb 2024
Project Page: https://doi.org/XXXXXXX.XXXXXXX
Paper: https://arxiv.org/pdf/2402.00485
