ACL RD-TEC 1.0 Summarization of P00-1061
Paper Title:
LEXICALIZED STOCHASTIC MODELING OF CONSTRAINT-BASED GRAMMARS USING LOG-LINEAR MEASURES AND EM TRAINING
LEXICALIZED STOCHASTIC MODELING OF CONSTRAINT-BASED GRAMMARS USING LOG-LINEAR MEASURES AND EM TRAINING
Authors: Stefan Riezler and Detlef Prescher and Jonas Kuhn and Mark Johnson
Primarily assigned technology terms:
- algorithm
- ambiguity resolution
- class-based grammar lexicalization
- classication
- clustering
- cross validation
- disambiguation
- em algorithm
- em training
- em-based clustering
- estimation method
- frame evaluation
- frame recognition
- iterative method
- iterative scaling
- language processing
- learner
- lexical ambiguity resolution
- lexicalization
- likelihood estimation
- likelihood maximization
- maximum entropy
- maximum entropy model
- maximum likelihood
- maximum likelihood estimation
- maximum pseudo-likelihood estimation
- modeling
- natural language processing
- parser
- parsers
- parsing
- preprocessing
- probabilistic modeling
- processing
- pseudo-likelihood estimation
- reading
- recognition
- search
- searching
- statistical disambiguation
- statistical inference
- statistical parsers
- statistical parsing
- subcategorization
- text classication
- training algorithm
- validation
Other assigned terms:
- accusative preposition
- adjunct
- adjunct attachment
- ambiguity
- ambiguity rate
- approach
- attribute-value pairs
- case
- class-based grammar
- clustering model
- conditional probability
- constraint-based grammar
- convergence
- corpora
- disambiguation task
- distribution
- entropy
- estimation
- evaluation measures
- exact match
- experimental results
- f-structure
- fact
- foreign language
- frame
- german newspaper text
- grammar
- grammar coverage
- grammars
- grammatical functions
- grammatical relation
- grammatical relations
- head word
- lexical ambiguity
- likelihood
- likelihood function
- linguistic
- linguistic environment
- linguistic phenomena
- local maxima
- log-likelihood
- log-linear model
- log-linear models
- main verb
- measure
- measures
- method
- natural language
- newspaper corpus
- nonterminal
- parameter values
- parse
- pcfg
- phrase
- precision
- preposition
- probabilities
- probability
- probability distribution
- procedure
- process
- recognition task
- relation
- search space
- sentence
- sentences
- statistics
- subcategorization frame
- subcorpus
- subtrees
- terms
- test corpora
- test corpus
- test data
- text
- theorem
- training
- training and test data
- training corpus
- training data
- unrestricted newspaper text
- verb
- verb frame
- verbal head
- word
- words
