A two-pass approach for handling out-of-vocabulary words in a large vocabulary recognition task.
Odette ScharenborgStephanie SeneffLou BovesPublished in: Comput. Speech Lang. (2007)
Keyphrases
- out of vocabulary
- continuous speech recognition
- n gram
- speech recognizer
- word segmentation
- language model
- speech recognition systems
- spoken document retrieval
- cross language information retrieval
- word recognition
- speech recognition
- broadcast news
- named entity recognition
- parallel corpora
- hand crafted
- cross lingual
- pattern recognition
- named entities
- term frequency
- language modeling
- word level
- query terms
- document analysis
- previously unseen
- test collection
- natural language processing
- knn
- information retrieval