This paper introduces a novel solution for personal recommendation in consumer electronic applications. It addresses, on the one hand, the data confidentiality during the training, by exploring federated learning and trusted ...