ACL RD-TEC 1.0 Summarization of P03-1021
Paper Title:
MINIMUM ERROR RATE TRAINING IN STATISTICAL MACHINE TRANSLATION
MINIMUM ERROR RATE TRAINING IN STATISTICAL MACHINE TRANSLATION
Primarily assigned technology terms:
- algorithm
- approximation
- automatic evaluation
- beam-search
- bootstrap
- bootstrap resampling
- chinese-english translation
- classification
- computing
- corresponding training
- direct translation
- dynamic programming
- dynamic programming beam-search
- dynamic programming search
- grid-based line optimization
- hypothesis translation
- language processing
- line optimization method
- loss function
- machine translation
- maximum entropy
- maximum likelihood
- modeling
- natural language processing
- optimization
- optimization algorithm
- optimization method
- parsing
- pattern recognition
- processing
- pruning
- recognition
- resampling
- scoring
- search
- search algorithm
- smoothing
- statistical machine translation
- subjective evaluation
- summarization
- training algorithm
- training procedure
- translation evaluation
- translation system
- word alignment
- words recognition
Other assigned terms:
- alignment template
- approach
- binary feature
- bleu
- bleu metric
- bleu score
- bleu scores
- brevity penalty
- candidate translation
- case
- classification error
- community
- convergence
- corpora
- corpus size
- correlation
- decision rule
- distribution
- document
- edit distance
- entropy
- entropy models
- error rate
- evaluation measure
- evaluation measures
- evaluation metric
- evaluation metrics
- fact
- feature
- french
- generation
- geometric mean
- grid
- human judgments
- human translation
- hypothesis
- language model
- likelihood
- log-linear model
- log-linear models
- machine translation quality
- maximum entropy models
- maximum entropy principle
- measure
- measures
- method
- model parameter
- model parameters
- modeling problem
- multi-reference word
- mutual information
- n-gram
- names
- natural language
- nist
- optimization criterion
- optimization problem
- parameter space
- parameter values
- phrase
- posterior
- posterior probability
- precision
- probabilistic model
- probabilistic models
- probability
- probability distribution
- probability distributions
- procedure
- reference translation
- reference translations
- relation
- reordering
- search problem
- sentence
- sentences
- smoothing parameter
- source sentence
- statistical models
- statistics
- symbol
- target language
- test corpora
- test corpus
- test data
- text
- training
- training corpus
- training criterion
- training data
- translation model
- translation models
- translation probability
- translation quality
- translation task
- translations
- true probability distribution
- word
- word error rate
- word order
- words
