ACL RD-TEC 1.0 Summarization of P04-1021
Paper Title:
A JOINT SOURCE-CHANNEL MODEL FOR MACHINE TRANSLITERATION
A JOINT SOURCE-CHANNEL MODEL FOR MACHINE TRANSLITERATION
Authors: Haizhou Li and Min Zhang and Jian Su
Primarily assigned technology terms:
- algorithm
- alignment process
- backoff smoothing
- bootstrap
- computational linguistics
- corpus alignment
- coupling
- cross-lingual information retrieval
- database
- databases
- decision tree
- decoder
- decoding
- em algorithm
- english-chinese transliteration
- error reduction
- finite state
- finite state transducer
- hidden markov
- hidden markov model
- id3 decision tree
- induction
- information retrieval
- language modeling
- learning
- learning algorithm
- learning algorithms
- learning approach
- learning techniques
- likelihood approach
- listing
- machine learning
- machine learning algorithms
- machine learning approach
- machine learning techniques
- machine translation
- machine transliteration
- markov model
- maximum likelihood
- maximum likelihood approach
- model training
- modeling
- name translation
- name transliteration
- noisy channel model
- noisy-channel model
- optimization
- orthographic mapping
- orthographical mapping
- reporting
- romanization
- search
- segmentation
- segmentation \/
- smoothing
- speech synthesis
- stack decoder
- statistical language modeling
- synthesis
- tokenization
- training process
- transducer
- transformation-based learning
- translator
- transliteration
- transliteration process
- tree construction
- viterbi
- viterbi algorithm
Other assigned terms:
- alphabet
- approach
- back-transliteration
- backoff
- bigram
- bilingual dictionary
- case
- character error rate
- character sequence
- characters
- chinese characters
- conditional probability
- contextual information
- data set
- data sets
- derivation
- dictionary
- distribution
- dom framework
- english-chinese language pair
- error rate
- exact match
- french
- implementation
- interpretation
- joint probability
- joint probability model
- katakana
- knowledge
- knowledge base
- language models
- language pair
- language pairs
- likelihood
- linguistics
- mapping
- mappings
- meaning
- meanings
- method
- multilingual corpus
- n-gram
- n-gram model
- n-gram models
- n-gram transliteration model
- names
- noisy channel
- open test
- orthography
- perplexity
- personal names
- phonemes
- phonemic representation
- pinyin
- precision
- probability
- probability distribution
- probability distributions
- probability model
- process
- proper names
- russian
- source language
- source-channel model
- speech synthesis literature
- statistics
- substring
- system development
- target language
- target languages
- test data
- test set
- tokens
- training
- training data
- training data set
- training database
- transformation
- transformation rules
- transliteration model
- tree
- tree path
- trees
- trigram
- word
- word error rates
- words
