Impact Analysis of the Use of Speech and Language Models Pretrained by Self-Supersivion for Spoken Language Understanding.
Salima MdhaffarValentin PelloinAntoine CaubrièreGaëlle LaperrièreSahar GhannayBassam JabaianNathalie CamelinYannick EstèvePublished in: LREC (2022)
Keyphrases
- language understanding
- language model
- impact analysis
- speech recognition
- spoken dialogue systems
- dialogue system
- word error rate
- language modeling
- dependency analysis
- natural language understanding
- source code
- n gram
- software development
- automatic speech recognition
- spoken term detection
- query expansion
- language processing
- life cycle
- speech signal
- probabilistic model
- retrieval model
- test collection
- semantic interpretation
- information retrieval
- signaling pathways
- natural language
- query terms
- general knowledge
- out of vocabulary
- human machine interaction
- vector space model
- smoothing methods
- context aware
- natural language processing
- low level