ACL RD-TEC 1.0 Summarization of H92-1021
Paper Title:
IMPROVEMENTS IN STOCHASTIC LANGUAGE MODELING
IMPROVEMENTS IN STOCHASTIC LANGUAGE MODELING
Authors: Ronald Rosenfeld and Xuedong Huang
Primarily assigned technology terms:
- approximation
- atis
- automatic speech recognition
- caching
- iterative reestimation
- language modeling
- language training
- linear interpolation
- modeling
- predictor
- processing
- pruning
- pruning method
- reasoning
- recognition
- recognition systems
- recognizer
- reestimation
- resource management
- smoothing
- smoothing method
- speech recognition
- speech recognition systems
- speech recognizer
- statistical analysis
- statistical reasoning
- stochastic language modeling
- word recognition
Other assigned terms:
- approach
- array
- backoff
- backoff language model
- backoff model
- bigram
- bigram model
- brown corpus
- cache
- case
- comprehension
- conditional probability
- contextual information
- corpora
- correlation
- correlations
- data set
- development set
- document
- events
- fact
- heuristics
- human reader
- index
- interpolation
- knowledge
- language model
- language models
- large corpus
- likelihood
- linguistic
- linguistic constraints
- location information
- measure
- measures
- method
- mutual information
- n-gram
- n-gram language model
- n-grams
- paragraph
- perplexity
- perplexity reduction
- probabilities
- probability
- probability estimate
- process
- recognition rate
- semantic
- sentence
- sentences
- sources of information
- test set
- testing data
- text
- training
- training and testing data
- training corpus
- training data
- training set
- trigram
- trigram model
- unigram
- unigram probability
- vocabulary
- word
- word sequence
- word sequences
- words
- wsj development set
