ACL RD-TEC 1.0 Summarization of P05-1022
Paper Title:
COARSE-TO-FINE N-BEST PARSING AND MAXENT DISCRIMINATIVE RERANKING
COARSE-TO-FINE N-BEST PARSING AND MAXENT DISCRIMINATIVE RERANKING
Authors: Eugene Charniak and Mark Johnson
Primarily assigned technology terms:
- algorithm
- beam search
- best-first parsing
- best-first search
- charniak parser
- collins parser
- conditional likelihood
- cross-validation
- discriminative reranking
- dynamic programming
- dynamic programming parsing
- estimation procedure
- estimator
- extractor
- feature selection
- hidden markov
- hidden markov models
- loss function
- maxent
- metric optimization
- n-best parsing
- optimization
- optimization algorithm
- optimization toolkit
- parser
- parsers
- parsing
- parsing algorithm
- probabilistic parser
- pruning
- regularization
- reranking
- right-branching
- scoring
- search
- search mechanism
- smoothing
- splitting
- statistical parsers
- variable metric optimization
- viterbi
- viterbi algorithm
Other assigned terms:
- approach
- auxiliary verb
- beam
- case
- conditional probability
- conditional probability model
- conjunct
- context-free grammar
- contextual features
- contextual information
- convergence
- data sets
- determiner
- development set
- estimation
- experimental results
- f-score
- fact
- feature
- feature vector
- feature weight
- feature weights
- forest
- grammar
- grammars
- heuristic
- implementation
- labeled training data
- lexical head
- lexical item
- lexical items
- likelihood
- local tree
- main verb
- maps
- markov models
- maxent conditional probability model
- measure
- method
- n-best list
- ngram
- nonterminal
- oracle
- packed parse forest
- parallelism
- parse
- parse forest
- parse tree
- parser output
- parsing model
- part of speech
- parts of speech
- pcfg
- pcfgs
- penn tree-bank
- penn treebank
- phrase
- probabilistic model
- probability
- probability model
- procedure
- projection
- punctuation
- root node
- rule schema
- schema
- sentence
- sentences
- subtrees
- technique
- term
- terms
- test data
- toolkit
- training
- training corpus
- training data
- tree
- tree-bank
- treebank
- trees
- user
- verb
- weight vector
- word
- words
