Effects of sub-word segmentation on performance of transformer language models.
Jue HouAnisia KatinskaiaAnh-Duc VuRoman YangarberPublished in: EMNLP (2023)
Keyphrases
- word segmentation
- language model
- language modeling
- n gram
- statistical language modeling
- word recognition
- information retrieval
- language independent
- document retrieval
- probabilistic model
- query expansion
- retrieval model
- handwriting recognition
- speech recognition
- test collection
- cross lingual
- out of vocabulary
- bag of words
- word level
- vector space model
- part of speech
- relevance model
- smoothing methods
- document level
- pseudo relevance feedback
- document ranking
- pattern recognition
- active learning
- machine learning