ACL RD-TEC 1.0 Summarization of H94-1015
Paper Title:
SPEECH RECOGNITION USING A STOCHASTIC LANGUAGE MODEL INTEGRATING LOCAL AND GLOBAL CONSTRAINTS
SPEECH RECOGNITION USING A STOCHASTIC LANGUAGE MODEL INTEGRATING LOCAL AND GLOBAL CONSTRAINTS
Authors: Ryosuke Isotani and Shoichi Matsunaga
Primarily assigned technology terms:
- algorithm
- approximation
- beam search
- binary decision tree
- candidate selection
- computing
- continuous speech recognition
- database
- decision tree
- decoding
- decomposition
- hmms
- interpolation method
- japanese speech recognition
- language modeling
- maximum likelihood
- modeling
- network representation
- normalization
- recognition
- recognition system
- recognizer
- rescoring
- search
- search process
- sentence recognition
- speaker-dependent speech recognition
- speech recognition
- speech recognition system
- speech recognizer
- spelling
- statistical approaches
- viterbi
- viterbi decoding
- weighting
- word bigram
Other assigned terms:
- acoustic score
- approach
- auxiliary verb
- auxiliary verbs
- beam
- bigram
- bigram model
- bunsetsu
- case
- case information
- content words
- continuous speech
- corpora
- distribution
- estimation
- fact
- function word
- function words
- hypotheses
- interpolation
- japanese sentences
- japanese text
- knowledge
- language model
- language models
- lattices
- likelihood
- linear combination
- linguistic
- linguistic constraints
- linguistic knowledge
- local constraints
- long distance dependencies
- markov models
- measure
- method
- model parameters
- mutual information
- n-gram
- n-gram model
- n-gram models
- n-grams
- nouns
- part of speech
- part of speech tags
- particles
- parts of speech
- perplexity
- phrase
- probabilities
- probability
- probability value
- process
- pronunciation
- recognition accuracy
- recognition rate
- recognition task
- semantic
- semantic constraints
- semantic information
- semantic relationships
- sentence
- sentences
- speech data
- speech recognition accuracy
- statistical measure
- statistics
- stochastic language model
- substring
- symbols
- syntactic and semantic information
- syntactic constraints
- tags
- telecommunications research
- term
- test data
- text
- text corpora
- text database
- training
- training and test data
- training data
- tree
- trigram
- trigram model
- uniform distribution
- verb
- vocabulary
- vocabulary size
- word
- word bigram model
- word pair
- word sequence
- word sequences
- word string
- word trigram
- word trigram model
- words
