Handling uncertain observations in unsupervised topic-mixture language model adaptation.
Ekapol ChuangsuwanichShinji WatanabeTakaaki HoriTomoharu IwataJames R. GlassPublished in: ICASSP (2012)
Keyphrases
- programming language
- decision making
- data driven
- natural language
- mixture model
- language learning
- unsupervised learning
- semi supervised
- topic modeling
- neural network
- possibilistic networks
- possibility theory
- computational linguistics
- language processing
- modeling language
- supervised learning
- gaussian mixture model
- gaussian distribution
- topic models
- expectation maximization
- probabilistic model
- knowledge base
- uncertain information
- unsupervised methods
- finite mixture model
- data sets