ACL RD-TEC 1.0 Summarization of W06-1638
Paper Title:
BETTER INFORMED TRAINING OF LATENT SYNTACTIC FEATURES
BETTER INFORMED TRAINING OF LATENT SYNTACTIC FEATURES
Authors: Markus Dreyer and Jason Eisner
Primarily assigned technology terms:
- add-lambda smoothing
- algorithm
- annealing clustering
- approximation
- bottom-up chart parsing
- capitalization
- chart parsing
- clustering
- clustering algorithm
- computational linguistics
- convex optimization
- decoding
- dynamic programming
- efficient parsing
- em algorithm
- em learning
- em training
- english parsing
- error analysis
- inside-outside algorithm
- language processing
- learner
- learning
- learning method
- learning procedure
- learning process
- log-linear modeling
- machine translation
- measuring
- modeling
- natural language processing
- optimization
- parser
- parsers
- parsing
- preprocessing
- processing
- reestimation
- reranking
- smoothing
- spelling
- splitting
- statistical machine translation
- statistical parsers
- supervised learning
- training algorithm
- unlexicalized parser
- unsupervised learning
- vertical markovization
- viterbi
Other assigned terms:
- annotation
- approach
- association for computational linguistics
- auxiliary verbs
- backbone grammar
- backoff
- base noun
- bias
- break
- capitalization information
- case
- citation
- cluster
- clusters
- conditional probabilities
- context-free grammar
- contextual information
- convergence
- corpora
- determiner
- development set
- distribution
- entropy
- fact
- feature
- feature value
- forest
- grammar
- grammars
- head noun
- heuristics
- ibm models
- implementation
- inheritance
- jensen-shannon divergence
- joint distribution
- lexical features
- lexical rule
- lexical rules
- likelihood
- linguistic
- linguistic constraints
- linguistic feature
- linguistic theories
- linguistics
- linguists
- log-linear model
- method
- model parameters
- model size
- mutual information
- natural language
- nonterminal
- nonterminals
- noun phrases
- optimization problem
- parameter values
- parse
- parse forest
- parse tree
- parsing accuracy
- part-of-speech
- part-of-speech tags
- pcfg
- pcfg-la model
- pcfgs
- penn treebank
- phrase
- precision
- prepositions
- probabilities
- probability
- procedure
- process
- rewrite rules
- runtime
- sentence
- sentences
- sparse data
- subject-verb agreement
- suffix
- symbol
- symbols
- syntactic features
- tags
- technique
- term
- terms
- test data
- test set
- theories
- training
- training corpora
- training corpus
- training data
- training time
- transformation
- tree
- treebank
- trees
- verb
- viterbi parse
- wall street journal corpus
- word
- words
