ACL RD-TEC 1.0 Summarization of H93-1023
Paper Title:
TOPIC AND SPEAKER IDENTIFICATION VIA LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION
TOPIC AND SPEAKER IDENTIFICATION VIA LARGE VOCABULARY CONTINUOUS SPEECH RECOGNITION
Authors: Barbara Peskin and Larry Gillick and Yoshiko Ito and Stephen Lowe and Robert Roth and Francesco Scattone and James Baker and Janet Baker and John Bridle and Melvyn Hunt and Jeremy Orloff
Primarily assigned technology terms:
- algorithm
- blind deconvolution
- classification
- classification system
- classifier
- classifiers
- computing
- continuous speech recognition
- continuous speech recognizer
- database
- databases
- deconvolution
- discriminant analysis
- editing
- factorization
- hidden markov
- hidden markov model
- identification
- identification system
- information retrieval
- information retrieval technology
- keyword selection
- language model training
- large vocabulary continuous speech recognition
- likelihood segmentation
- linear discriminant
- linear discriminant analysis
- markov model
- maximum likelihood
- maximum likelihood segmentation
- message classification
- message processing
- model training
- modelling
- processing
- recognition
- recognition system
- recognizer
- retrieval technology
- scoring
- searching
- segmentation
- signal-processing
- smoothing
- speaker identification
- speaker-independent recognition
- speech recognition
- speech recognition system
- speech recognizer
- stochastic process
- text-based information retrieval
- topic classification
- topic classifier
- topic identification
- topic scoring
- topic-independent recognizer
- transcription
- vocabulary recognition
- word recognition
- wordspotting
Other assigned terms:
- acoustic model
- acoustic models
- approach
- backoff
- binomial distribution
- case
- classification problem
- conditional probability
- contextual information
- continuous speech
- conversational speech
- corpora
- data consortium
- distribution
- duration
- frame
- function words
- hypotheses
- hypothesis
- identification task
- keyword
- labeling
- language model
- language models
- likelihood
- linguistic
- linguistic data
- linguistic data consortium
- mapping
- message
- method
- model context
- noise
- null hypothesis
- parameter space
- phoneme
- phonemes
- preprocessor
- prior probability
- probabilities
- probability
- probability model
- process
- pronunciation
- recognition task
- scoring model
- segments
- silence
- speech data
- statistic
- stochastic model
- switchboard corpus
- technology
- term
- test material
- test set
- text
- textual corpora
- theoretical framework
- theory
- topic identification task
- topics
- training
- training corpus
- training data
- training material
- training text
- transcribed switchboard
- transcript
- transcriptions
- transcripts
- transformation
- unigram
- vocabulary
- word
- word frequencies
- word frequency
- word pronunciation
- word sequence
- word sequences
- word strings
- words
