Joint unsupervised adaptation of n-gram and RNN language models via LDA-based hybrid mixture modeling.
Ryo MasumuraTaichi AsamiHirokazu MasatakiYushi AonoPublished in: APSIPA (2017)
Keyphrases
- language model
- n gram
- mixture modeling
- mixture model
- language modeling
- density estimation
- probabilistic model
- language modelling
- information retrieval
- document retrieval
- language independent
- unsupervised learning
- nearest neighbor
- retrieval model
- query expansion
- em algorithm
- part of speech
- test collection
- vector space model
- query terms
- pseudo relevance feedback
- relevance model
- statistical language modeling
- data mining
- semi supervised
- word segmentation
- dirichlet distribution
- naive bayes
- expectation maximization
- supervised learning
- out of vocabulary
- training set
- decision trees