ACL RD-TEC 1.0 Summarization of N03-1027
Paper Title:
SUPERVISED AND UNSUPERVISED PCFG ADAPTATION TO NOVEL DOMAINS
SUPERVISED AND UNSUPERVISED PCFG ADAPTATION TO NOVEL DOMAINS
Authors: Brian Roark and Michiel Bacchiani
Primarily assigned technology terms:
- acoustic modeling
- active learning
- algorithm
- automatic speech recognition
- beam search
- beam-search
- chunking
- expectation maximization
- expectation maximization algorithm
- frequency estimation
- grammar induction
- incremental parser
- induction
- inside-outside algorithm
- language modeling
- language modeling approach
- language processing
- learning
- learning techniques
- lexicalized parser
- likelihood estimate
- map adaptation
- map estimation
- maximization algorithm
- maximum entropy
- maximum entropy model
- maximum likelihood
- model adaptation
- model interpolation
- model mixing
- modeling
- n-gram modeling
- optimization
- parameter optimization
- parameter tuning
- parameter tying
- parameterization
- parser
- parsers
- parsing
- parsing algorithm
- pos tagging
- posteriori estimation
- pre-processing
- processing
- re-training
- recognition
- recognizer
- scoring
- search
- smoothing
- speech recognition
- speech recognizer
- statistical estimation
- statistical parsers
- statistical parsing
- taggers
- tagging
- top-down parser
- tuning
- weighting
Other assigned terms:
- adaptation schema
- annotated corpora
- annotated corpus
- annotated treebank
- annotation
- approach
- baseline model
- baseline performance
- beam
- brown corpus
- case
- conditional probabilities
- conditional probability
- context-free grammar
- context-free grammars
- corpora
- dirichlet distribution
- distribution
- empirical results
- entropy
- estimation
- f-measure
- fact
- feature
- grammar
- grammars
- heuristic
- interpolation
- joint probability
- labeled training data
- language models
- likelihood
- manual annotation
- maximum likelihood estimate
- model parameters
- n-gram
- n-gram model
- n-gram models
- optimization problem
- parse
- parser output
- parser performance
- parsing accuracy
- parsing model
- parsing models
- part-of-speech
- partial parse
- parts-of-speech
- pcfg
- pcfg model
- pcfgs
- penn treebank
- penn treebank ii
- penn wsj treebank
- posterior
- posterior distribution
- precision
- prior distribution
- probabilistic context-free grammars
- probabilistic model
- probabilities
- probability
- probability estimate
- procedure
- rapid adaptation
- relative frequency
- right-hand side
- schema
- sentences
- statistic
- statistical model
- symbol
- symbols
- terms
- test set
- training
- training corpus
- training data
- training set
- transformation
- treebank
- trees
- word
- words
- wsj treebank
