A novel privacy-preserving matrix factorization recommendation system based on random perturbation.
Zhaoyan HuYonglong LuoXiaoyao ZhengYannian ZhaoPublished in: J. Intell. Fuzzy Syst. (2020)
Keyphrases
- privacy preserving
- matrix factorization
- collaborative filtering
- data perturbation
- recommender systems
- privacy preserving data mining
- privacy guarantees
- low rank
- vertically partitioned data
- privacy preservation
- factorization methods
- private information
- multi party
- data privacy
- nonnegative matrix factorization
- negative matrix factorization
- recommendation systems
- private data
- missing data
- privacy sensitive
- differential privacy
- sensitive information
- preserving privacy
- horizontally partitioned data
- privacy preserving association rule mining
- privacy concerns
- secure multiparty computation
- support vector
- user preferences
- linear combination
- homomorphic encryption
- scalar product
- partitioned data