A differentially private matrix factorization based on vector perturbation for recommender system.
Xun RanYong WangLeo Yu ZhangJun MaPublished in: Neurocomputing (2022)
Keyphrases
- matrix factorization
- recommender systems
- differentially private
- collaborative filtering
- differential privacy
- privacy guarantees
- low rank
- negative matrix factorization
- user preferences
- nonnegative matrix factorization
- factorization methods
- cold start problem
- data sparsity
- probabilistic matrix factorization
- user model
- implicit feedback
- cold start
- item recommendation
- user profiles
- missing data
- contingency tables
- latent factors
- privacy preserving
- data analysis
- data mining