Federated against the cold: A trust-based federated learning approach to counter the cold start problem in recommendation systems

Omar Abdel Wahab, Gaith Rjoub, Jamal Bentahar, Robin Cohen

Abstract: Highlights•Federated learning-based recommendation system for cold-start items.•Trust establishment for recommenders that considers resource utilization and credibility.•Recommender selection strategy based on Double Deep Q Learning.•Simulations on the MovieLens 1M dataset suggest better accuracy compared to two benchmark approaches.