ACL RD-TEC 1.0 Summarization of W04-3255
Paper Title:
EFFICIENT DECODING FOR STATISTICAL MACHINE TRANSLATION WITH A FULLY EXPANDED WFST MODEL
EFFICIENT DECODING FOR STATISTICAL MACHINE TRANSLATION WITH A FULLY EXPANDED WFST MODEL
Authors: Hajime Tsukada and Masaaki Nagata
Primarily assigned technology terms:
- algorithm
- ambiguity reduction
- beam search
- beam-search
- breadth-first search
- computing
- decoder
- decoding
- depth-first search
- finite-state acceptor
- finite-state transducer
- forward beam-search
- forward pass
- forward search
- giza
- hmms
- language modeling
- machine translation
- machine translation system
- modeling
- optimization
- optimization method
- phrase-based translation
- pruning
- recognition
- rule-based machine
- rule-based machine translation
- search
- search method
- search process
- searching
- speech recognition
- state merging
- statistical language modeling
- statistical machine translation
- transducer
- translation process
- translation system
- weighted finite-state transducer
- wfst optimization
- word alignment
Other assigned terms:
- ambiguity
- approach
- back-off parameter
- beam
- bilingual corpora
- bilingual corpus
- bleu
- bleu score
- composition
- conditional probability
- corpora
- device
- distortion probability
- english sentence
- english vocabulary
- experimental results
- fact
- hypotheses
- hypothesis
- ibm model
- ibm models
- implementation
- japanese sentences
- japanese word
- language model
- language modeling toolkit
- measure
- method
- model size
- modeling toolkit
- n-gram
- nist
- noisy channel
- permutation
- probability
- process
- sentence
- sentences
- source language
- statistic
- statistical model
- statistical models
- statistics
- symbols
- target language
- technique
- text
- theory
- toolkit
- training
- training set
- translation accuracy
- translation model
- translation probability
- translation problem
- translations
- trigram
- vocabulary
- word
- word alignments
- word trigram
- words
