This initiative, led by a researcher under the supervision of a professor, positions the UJA at the international forefront of research in responsible AI and group recommendation systems.
The Fidelity project addresses the challenge of combining fairness, explainability, and trust in AI systems that influence daily decisions, from e-commerce platforms to news and digital service recommendations.
“"The project has developed new algorithms and tools capable not only of generating high-quality recommendations but also of explaining why they are suitable and how they ensure balanced treatment for users and providers."
One of the key advances has been demonstrating that explainability and algorithmic fairness are complementary dimensions for building trustworthy systems. Researchers have validated their proposals through experiments and user studies, analyzing the perception of transparency and satisfaction.
The relevance of these results has been recognized in prestigious scientific forums, such as ACM UMAP, where studies derived from the project have empirically validated the impact of these new explainable and fair strategies in group recommendation. Additionally, contributions have been made to international conferences and journals on explanation and fairness in recommendation platforms.
Beyond the scientific contribution, Fidelity offers a practical roadmap for developers and technology companies to integrate responsible AI principles into real products, fostering more balanced and understandable digital ecosystems for users and providers.
