Recommending with an agenda: active learning of private attributes using matrix factorization.
Smriti BhagatUdi WeinsbergStratis IoannidisNina TaftPublished in: RecSys (2014)
Keyphrases
- matrix factorization
- active learning
- collaborative filtering
- recommender systems
- factorization methods
- low rank
- missing data
- data sparsity
- negative matrix factorization
- user preferences
- data matrix
- probabilistic matrix factorization
- semi supervised
- factor analysis
- latent factors
- nonnegative matrix factorization
- transfer learning
- variational bayesian
- item recommendation
- relevance feedback
- machine learning
- personalized recommendation
- user ratings
- stochastic gradient descent
- learning algorithm
- random sampling
- implicit feedback
- recommendation systems
- latent factor models
- user feedback
- binary matrix
- user profiles
- factor matrices
- labeled data