ACL RD-TEC 1.0 Summarization of I05-2021
Paper Title:
EVALUATING THE WORD SENSE DISAMBIGUATION PERFORMANCE OF STATISTICAL MACHINE TRANSLATION
EVALUATING THE WORD SENSE DISAMBIGUATION PERFORMANCE OF STATISTICAL MACHINE TRANSLATION
Authors: Marine CARPUAT and Dekai WU
Primarily assigned technology terms:
- algorithm
- bayes classifier
- bilingual bootstrapping
- boosting
- bootstrapping
- bootstrapping method
- classification
- classifier
- classifier ensemble
- classifiers
- computing
- decoder
- decoding
- disambiguation
- error analysis
- evaluation framework
- giza
- greedy decoding
- information retrieval
- isi rewrite decoder
- kernel
- language modeling
- language processing
- machine translation
- machine translation system
- maximum entropy
- maximum entropy model
- modeling
- mt system
- mt systems
- naive bayes
- naive bayes classifier
- natural evaluation
- natural language processing
- parallel training
- processing
- sense disambiguation
- smt system
- statistical language modeling
- statistical machine translation
- statistical machine translation system
- statistical mt
- supervised training
- translation disambiguation
- translation system
- unsupervised approach
- unsupervised bootstrapping
- unsupervised wsd
- word alignment
- word level disambiguation
- word sense disambiguation
- word translation
Other assigned terms:
- alignment model
- alignment models
- ambiguity
- annotated corpus
- annotation
- approach
- bayes model
- bias
- bleu
- bleu score
- bleu scores
- characters
- chinese lexical
- chinese sentence
- chinese word
- classification task
- cohesion
- comparable corpora
- contextual information
- corpora
- data sets
- disambiguation model
- empirical results
- english language
- english language model
- english translations
- entropy
- error rate
- evaluation data
- evaluation methodology
- evaluation metrics
- evaluation set
- evaluation test
- fact
- feature
- feature set
- grammaticality
- hindi
- hownet
- knowledge
- knowledge base
- language model
- language modeling toolkit
- language processing applications
- lexical choice
- linguistic
- linguistic information
- local context
- long distance dependencies
- machine translation quality
- manual annotation
- mapping
- method
- methodology
- modeling toolkit
- mt evaluation
- n-grams
- naive bayes model
- natural language
- natural language processing applications
- parallel corpora
- parallel corpus
- parallel training corpus
- phrase
- probabilities
- sense ambiguity
- sense disambiguation model
- sense distinctions
- sense inventory
- senses of a word
- sentence
- sentences
- sentential context
- smt translation
- source language
- surface form
- target language
- target sentence
- target word
- terms
- test set
- toolkit
- training
- training corpus
- training data
- training instance
- training set
- translation candidates
- translation model
- translation models
- translation probabilities
- translation quality
- translation task
- translations
- tree
- trigram
- trigram model
- word
- word error rate
- word level
- word sense
- word type
- word types
- wordnet
- wordnet senses
- words
- wsd evaluation
- wsd evaluation methodology
- wsd model
