N-Grams for Conditional Random Fields or a Failure-Transition(f) Posterior for Acyclic FSTs.
Patrick LehnenStefan HahnHermann NeyPublished in: INTERSPEECH (2011)
Keyphrases
- conditional random fields
- n gram
- probabilistic model
- language model
- graphical models
- hidden markov models
- text classification
- language modeling
- sequence labeling
- generative model
- pairwise
- markov random field
- part of speech
- information extraction
- higher order
- bag of words
- variable length
- random fields
- maximum entropy
- markov networks
- structured prediction
- semi markov
- posterior probability
- crf model
- bayesian framework
- probability distribution
- character n grams
- web page prediction
- named entity recognition
- multi class
- feature selection
- neural network