A hybrid approach of topic model and matrix factorization based on two-step recommendation framework.
Xiangyu ZhaoZhendong NiuWei ChenChongyang ShiKe NiuDonglei LiuPublished in: J. Intell. Inf. Syst. (2015)
Keyphrases
- matrix factorization
- topic models
- latent factor models
- collaborative filtering
- recommender systems
- topic modeling
- latent factors
- latent dirichlet allocation
- item recommendation
- probabilistic model
- text mining
- recommendation systems
- user feedback
- data sparsity
- negative matrix factorization
- factor analysis
- multi task
- transfer learning
- semi supervised
- active learning