Efficient Multilingual Language Model Compression through Vocabulary Trimming.
Asahi UshioYi ZhouJosé Camacho-ColladosPublished in: EMNLP (Findings) (2023)
Keyphrases
- language model
- language modeling
- n gram
- document retrieval
- cross lingual
- speech recognition
- probabilistic model
- query expansion
- information retrieval
- language modelling
- language independent
- mixture model
- retrieval model
- context sensitive
- query specific
- statistical language models
- out of vocabulary
- smoothing methods
- test collection
- relevance model
- pseudo relevance feedback
- text classification
- hidden markov models