ACL RD-TEC 1.0 Summarization of P05-1010
Paper Title:
PROBABILISTIC CFG WITH LATENT ANNOTATIONS
PROBABILISTIC CFG WITH LATENT ANNOTATIONS
Authors: Takuya Matsuzaki and Yusuke Miyao and Jun’ichi Tsujii
Primarily assigned technology terms:
- algorithm
- approximation
- beam thresholding
- cky algorithm
- constrained optimization
- dynamic programming
- dynamic programming algorithm
- em algorithm
- estimation algorithm
- exact inference
- exact parsing
- feature selection
- hmms
- inside-outside algorithm
- inside-outside re-estimation
- lagrange multiplier
- neural network
- optimization
- parameter estimation
- parser
- parsers
- parsing
- pcfg parser
- pcfg parsing
- pre-processing
- programming algorithm
- re-estimation
- re-ranking
- reading
- search
- subcategorization
- thresholding
- tree substitution
- viterbi
- viterbi algorithm
Other assigned terms:
- annotated corpora
- annotation
- approach
- beam
- case
- complete parse
- conditional independence
- conditional probability
- corpora
- dependency relations
- derivation
- derivations
- development set
- distribution
- empty nodes
- estimation
- evaluation metric
- fact
- feature
- forest
- generation
- generative model
- grammar
- grammar rules
- independence assumption
- likelihood
- measure
- method
- model size
- non-terminal symbol
- optimization problem
- parse
- parse time
- parse tree
- parsed corpus
- parsing model
- pcfg
- pcfg model
- pcfg-la model
- pcfgs
- polynomial time
- posterior
- posterior probability
- probabilistic model
- probabilities
- probability
- process
- root node
- sentence
- sentences
- stem
- subcategorization frames
- substitution grammar
- symbol
- symbols
- tags
- terms
- test data
- test set
- theory
- time complexity
- trained model
- training
- training data
- training set
- training time
- tree
- tree substitution grammar
- treebank
- trees
- viterbi parse
- word
- words
- wsj corpus
