ACL RD-TEC 1.0 Summarization of N01-1017
Paper Title:
GENERATING TRAINING DATA FOR MEDICAL DICTATIONS
GENERATING TRAINING DATA FOR MEDICAL DICTATIONS
Authors: Sergey Pakhomov and Michael Schonwetter and Joan Bachenko
Primarily assigned technology terms:
- acoustic adaptation
- acoustic adaptation training
- acoustic model adaptation
- acoustic modeling
- approximation
- automatic speech recognition
- data representation
- dynamic programming
- dynamic programming alignment
- finite state
- language model adaptation
- language model training
- language modeling
- linear interpolation
- matching
- measuring
- medical dictation
- model adaptation
- model training
- modeling
- post-processing
- recognition
- recognizer
- reformatting
- si modeling
- speech recognition
- spoken discourse
- transcription
- transcription reconstruction
- transcription reconstruction system
Other assigned terms:
- acoustic model
- acoustic models
- acoustic signal
- background model
- background noise
- confidence scores
- corpora
- dictionary
- discourse
- experimental results
- fact
- finite state model
- formalism
- frame
- generation
- grammars
- hypothesis
- interpolation
- interpreter
- language model
- language model perplexity
- language model quality
- language models
- large corpus
- meaning
- measure
- measures
- method
- model perplexity
- natural speech
- noise
- nouns
- paragraph
- paragraphs
- patent
- pauses
- perplexity
- prepositions
- probabilities
- process
- punctuation
- recognition accuracy
- segments
- sentence
- signal
- speech signal
- statistical significance
- technique
- testing data
- text
- toolkit
- training
- training corpus
- training data
- training set
- training text
- transcript
- transcriptions
- transcripts
- transformation
- trigram
- utterance
- vocabulary
- word
- word boundaries
- words
