ACL RD-TEC 1.0 Summarization of C04-1014
Paper Title:
MODELING OF LONG DISTANCE CONTEXT DEPENDENCY
MODELING OF LONG DISTANCE CONTEXT DEPENDENCY
Primarily assigned technology terms:
- character recognition
- chinese character recognition
- computing
- hmms
- language modeling
- language processing
- likelihood estimation
- mandarin speech recognition
- maximum likelihood
- maximum likelihood estimation
- mi-ngram modeling
- modeling
- natural language processing
- ngram modeling
- processing
- recognition
- recognition system
- sampling
- smoothing
- smoothing technique
- speech recognition
- speech recognition system
- statistical language modeling
- syllable recognition
- word bigram
- word ngram modeling
Other assigned terms:
- approach
- bigram
- bigram model
- characters
- chinese characters
- concept
- conditional distribution
- context dependency
- data sparseness
- data sparseness problem
- distribution
- entropy
- estimation
- feature
- feature vector
- hypothesis
- joint probability
- lexicon
- likelihood
- mandarin chinese
- measure
- modeling power
- mutual information
- natural language
- news corpus
- ngram
- ngram model
- perplexity
- pitch
- probabilities
- probability
- recognition rate
- sentences
- sparseness problem
- statistical models
- technique
- testing data
- text
- tone
- training
- training and testing data
- training text
- trigram
- trigram model
- window size
- word
- word pair
- word string
- word strings
- word trigram
- word trigram model
- words
- xinhua corpus