Document Hashing with Mixture-Prior Generative Models.
Wei DongQinliang SuDinghan ShenChangyou ChenPublished in: EMNLP/IJCNLP (1) (2019)
Keyphrases
- generative model
- mixture model
- prior knowledge
- dirichlet distribution
- expectation maximization
- probabilistic model
- bayesian framework
- dirichlet process
- em algorithm
- discriminative learning
- latent dirichlet allocation
- document clustering
- document retrieval
- hierarchical hidden markov models
- conditional random fields
- information retrieval
- maximum entropy principle
- document collections
- document classification
- text documents
- lda model
- information retrieval systems
- discriminative models
- object categories
- retrieval systems
- hidden variables
- semi supervised
- topic models
- generative and discriminative models
- pairwise
- training data
- representational power