FedeRank: User Controlled Feedback with Federated Recommender Systems

Recommender systems have shown to be a successful representative of how data availability can ease our everyday digital life. However, privacy is one the most prominent concerns in era. After several breaches and scandals, users are now worried about sharing their data. In last decade, Federated Learning has emerged as new privacy-preserving distributed machine learning paradigm. It works by processing on user device without collecting central repository. We present FedeRank ( https://split.to/federank ), federated recommendation algorithm. The system learns personal factorization model onto every device. training synchronous process between server clients. takes care computing recommendations fashion allows control portion they want share. By comparing with state-of-the-art algorithms, extensive experiments show effectiveness terms accuracy, even small shared Further analysis lists’ diversity novelty guarantees suitability algorithm real production environments.