Optimizing CRFs for SLU tasks in various languages using modified training criteria.
Stefan HahnPatrick LehnenGeorg HeigoldHermann NeyPublished in: INTERSPEECH (2009)
Keyphrases
- conditional random fields
- sequence labeling
- structured prediction
- databases
- training algorithm
- training set
- information extraction
- training process
- named entity recognition
- expressive power
- probabilistic model
- test set
- training examples
- higher order
- evaluation criteria
- supervised learning
- multi criteria
- text summarization
- training phase
- parameter learning
- selection criteria
- error correcting output coding
- dependency parsing
- language independent
- maximum likelihood
- pairwise
- similarity measure