Quand être absent de mBERT n'est que le commencement : Gérer de nouvelles langues à l'aide de modèles de langues multilingues (When Being Unseen from mBERT is just the Beginning : Handling New Languages With Multilingual Language Models).
Benjamin MüllerAntonios AnastasopoulosBenoît SagotDjamé SeddahPublished in: TALN-RECITAL (2022)
Keyphrases
- language model
- language modeling
- cross lingual
- language independent
- n gram
- comparable corpora
- cross lingual information retrieval
- translation model
- cross language
- probabilistic model
- multilingual information retrieval
- statistical machine translation
- retrieval model
- document retrieval
- language modelling
- query expansion
- context sensitive
- speech recognition
- statistical language models
- information retrieval
- query specific
- language model for information retrieval
- test collection
- machine translation system
- machine translation
- smoothing methods
- relevance model
- parallel corpora
- term dependencies
- linguistic resources
- word segmentation
- chinese english
- text classification
- training set
- document ranking
- document level
- query terms
- web search
- hidden markov models
- language models for information retrieval