The Source-Target Domain Mismatch Problem in Machine Translation.
Jiajun ShenPeng-Jen ChenMatt LeJunxian HeJiatao GuMyle OttMichael AuliMarc'Aurelio RanzatoPublished in: CoRR (2019)
Keyphrases
- machine translation
- target domain
- source domain
- domain adaptation
- transfer learning
- cross domain
- multiple sources
- labeled data
- target language
- training data
- domain specific
- cross lingual
- test data
- training samples
- cross language information retrieval
- unlabeled data
- natural language processing
- information extraction
- source language
- statistical machine translation
- labeled training data
- word alignment
- active learning
- text categorization
- semi supervised learning
- semi supervised
- query translation
- data distribution
- label information
- sentiment classification
- knowledge transfer
- data sets
- text classification
- text mining
- supervised learning
- data points
- high dimensional
- learning algorithm
- information retrieval