Dual Conditional Cross-Entropy Filtering of Noisy Parallel Corpora.
Marcin Junczys-DowmuntPublished in: WMT (shared task) (2018)
Keyphrases
- cross entropy
- parallel corpora
- cross lingual
- language modeling
- language independent
- cross language information retrieval
- machine translation
- maximum likelihood
- log likelihood
- labor intensive
- machine translation system
- statistical machine translation
- word pairs
- cross language
- evaluation metrics
- query translation
- error function
- multimedia
- feature space
- sentence level
- wikipedia articles
- search engine
- error prone