ACL RD-TEC 1.0 Summarization of P97-1063
Paper Title:
A WORD-TO-WORD MODEL OF TRANSLATIONAL EQUIVALENCE
A WORD-TO-WORD MODEL OF TRANSLATIONAL EQUIVALENCE
Primarily assigned technology terms:
- algorithm
- bilingual concordancer
- broad-coverage translation
- classification
- cutoff
- direct association
- em algorithm
- estimation method
- hill-climbing
- induction
- induction algorithm
- likelihood estimation
- linking
- matching
- maximum likelihood
- maximum likelihood estimation
- model induction
- modeling
- nlp
- parameter estimation
- part-of-speech tagger
- probability function
- statistical translation
- table lookup
- tagger
- thresholding
- tokenization
- translation model induction
- word-to-word translation
Other assigned terms:
- aligned sentence
- auxiliary verbs
- bilingual dictionaries
- binomial distribution
- bitext
- bitext space
- candidate translation
- case
- co-occurrence
- co-occurrence frequency
- co-occurrence relation
- co-occurrences
- coefficient
- content words
- convergence
- dice
- dice coefficient
- dictionaries
- dictionary
- distribution
- entropy
- estimation
- feature
- french
- french word
- frequency distribution
- function words
- heuristic
- ibm model
- ibm models
- knowledge
- lexicon
- lexicon entry
- likelihood
- likelihood ratio
- local maxima
- log-likelihood
- measure
- measures
- method
- model parameters
- nlp applications
- parallel texts
- parameter space
- paraphrase
- paraphrases
- part of speech
- part-of-speech
- partof-speech
- parts of speech
- precision
- preprocessor
- priori
- probabilities
- probability
- relation
- representations
- right-hand side
- running time
- segments
- sentence
- sentences
- stems
- sun microsystems
- terms
- test set
- text
- tokens
- training
- training data
- translation lexicon
- translation model
- translation models
- translational equivalence
- translations
- vocabulary
- word
- word co-occurrence
- word level
- word order
- word type
- word types
- word-level model
- word-to-word translation model
- words
