Privacy-Preserving Matrix Factorization for Cross-Domain Recommendation

Cross-domain recommender systems are known to provide solutions the cold start and data sparsity problems in systems. This can be achieved by leveraging sufficient ratings users' profiles one domain enhance accurate recommendations another domain. However, domains with not willing share their other or due privacy legal concern. Hence this shows a need for privacy-preserving mechanism that encourages secure knowledge transfer between different domains. study proposes cross-domain system based on matrix factorization. Specifically, formally described requirements of system, which from single system. It designs new framework then utilized somewhat homomorphic encryption (SWHE) scheme ensure privacy. The SWHE was used encrypt domains, shared latent factor approach implemented extracted securely transferred source target We prove is secured throughout stages involved proposed protocol. Experiments both synthetic real datasets demonstrate efficiency our