Continual Mixed-Language Pre-Training for Extremely Low-Resource Neural Machine Translation.
Zihan LiuGenta Indra WinataPascale FungPublished in: CoRR (2021)
Keyphrases
- machine translation
- target language
- language processing
- machine translation system
- language specific
- language resources
- source language
- natural language
- parallel corpus
- cross language information retrieval
- multilingual documents
- language independent
- cross lingual
- natural language processing
- statistical machine translation
- word level
- information extraction
- comparable corpora
- word alignment
- phrase based smt
- word sense disambiguation
- chinese english
- bilingual dictionaries
- brazilian portuguese
- linguistic knowledge
- natural language generation
- parallel corpora
- tasks in natural language processing
- bilingual lexicon
- finite state transducers
- pos tagging
- knowledge representation