ACL RD-TEC 1.0 Summarization of C96-2154
Paper Title:
MODELING TOPIC COHERENCE FOR SPEECH RECOGNITION
MODELING TOPIC COHERENCE FOR SPEECH RECOGNITION
Primarily assigned technology terms:
- continuous speech recognition
- information retrieval
- information retrieval systems
- keyword selection
- machine translation
- machine translation system
- modeling
- optimization
- optimization procedure
- parameter optimization
- recognition
- recognition system
- recognizer
- retrieval method
- retrieval process
- retrieval systems
- retrieving
- similarity calculation
- speech recognition
- speech recognition evaluation
- speech recognition system
- speech recognizer
- sublanguage analysis
- tile
- translation system
- weight optimization
Other assigned terms:
- approach
- automatic optimization
- cache
- case
- coherence
- continuous speech
- data set
- discourse
- document
- document sets
- error rate
- evaluation data
- fact
- function word
- hypotheses
- hypothesis
- information retrieval research
- keyword
- language model
- language model score
- language models
- large corpus
- linear combination
- method
- nist
- noise
- parameter settings
- parameter values
- part of speech
- perplexity
- probability
- procedure
- process
- recognition evaluation
- sentence
- sentence level
- sentences
- speech data
- statistic
- statistical language model
- statistical model
- sublanguage
- systeln
- target language
- technique
- text
- text corpus
- tile corpus
- tile evaluation
- tile system
- tokens
- topic continuity
- topics
- training
- training corpus
- training data
- training data set
- training set
- transcriptions
- trigram
- trigram model
- unigram
- unigram probability
- word
- word choice
- word error rate
- word level
- word trigram
- word trigram model
- words
