On the Weaknesses of Reinforcement Learning for Neural Machine Translation.
Leshem ChoshenLior FoxZohar AizenbudOmri AbendPublished in: CoRR (2019)
Keyphrases
- machine translation
- reinforcement learning
- language independent
- natural language processing
- information extraction
- language processing
- cross language information retrieval
- cross lingual
- chinese english
- statistical machine translation
- natural language
- target language
- language resources
- word sense disambiguation
- parallel corpora
- machine translation system
- query translation
- natural language generation
- state space
- markov decision processes
- optimal policy
- dynamic programming
- learning algorithm
- multilingual documents
- word level
- precision and recall
- artificial intelligence