ACL RD-TEC 1.0 Summarization of P06-1065
Paper Title:
IMPROVED DISCRIMINATIVE BILINGUAL WORD ALIGNMENT
IMPROVED DISCRIMINATIVE BILINGUAL WORD ALIGNMENT
Authors: Robert C. Moore and Wen-tau Yih and Andreas Bode
Primarily assigned technology terms:
- 5-fold cross validation
- algorithm
- alignment combination
- automatic sentence alignment
- beam search
- bilingual word alignment
- computational linguistics
- cross validation
- cross-validation
- cross-validation method
- decoding
- discriminative training
- five-fold cross validation
- giza
- large-margin training
- large-margin training technique
- learner
- learning
- learning methods
- likelihood estimate
- machine translation
- maximum likelihood
- measuring
- parser
- perceptron
- perceptron learning
- perceptron training
- principal training
- search
- sentence alignment
- smoothing
- statistical machine translation
- structured learning
- support vector machine
- svm learning
- training method
- training procedure
- training process
- tuning
- validation
- word alignment
- word-alignment
Other assigned terms:
- alignment error rate
- alignment model
- alignment models
- approach
- association for computational linguistics
- association score
- beam
- beam size
- bilingual corpus
- bilingual dictionary
- bilingual sentence
- case
- cluster
- clusters
- conditional probabilities
- conditional probability
- convergence
- dictionary
- discriminative model
- error rate
- exact match
- fact
- feature
- feature value
- feature weight
- feature weights
- function word
- generative models
- hard constraint
- heuristic
- hmm model
- ibm model
- ibm models
- labeled training data
- learning rate
- lexical features
- likelihood
- likelihood ratio
- linear combination
- linguistics
- maximum likelihood estimate
- measure
- method
- model structure
- names
- parameter values
- precision
- probabilities
- probability
- procedure
- process
- proper names
- search procedure
- sentence
- sentence pair
- sentences
- source sentence
- statistic
- statistics
- support vector
- target sentence
- technique
- terms
- test data
- test set
- tokens
- training
- training and test data
- training corpus
- training data
- training example
- training examples
- training set
- weight vector
- word
- word alignment model
- word alignments
- word association
- word order
- words
