ACL RD-TEC 1.0 Summarization of W02-1031
Paper Title:
THE SUPERARV LANGUAGE MODEL: INVESTIGATING THE EFFECTIVENESS OF TIGHTLY INTEGRATING MULTIPLE KNOWLEDGE SOURCES
THE SUPERARV LANGUAGE MODEL: INVESTIGATING THE EFFECTIVENESS OF TIGHTLY INTEGRATING MULTIPLE KNOWLEDGE SOURCES
Authors: Wen Wang and Mary P. Harper
Primarily assigned technology terms:
- algorithm
- bracketing
- computational linguistics
- constraint relaxation
- context-free grammar parser
- continuous speech recognition
- coupling
- decoder
- feature augmentation
- grammar parser
- grouping
- kneser-ney smoothing
- language modeling
- language processing
- lattice rescoring
- learning
- likelihood estimation
- linear interpolation
- link grammar
- maximum likelihood
- maximum likelihood estimation
- modeling
- parameter reestimation
- parameter tuning
- parser
- parsing
- preprocessing
- processing
- pruning
- ranking
- recognition
- recognition systems
- recognizer
- reestimation
- rescoring
- rule learning
- rule-based method
- search
- smoothing
- speech recognition
- speech recognition systems
- speech recognizer
- subcategorization
- tokenization
- tuning
- viterbi
- viterbi search
- word prediction
- wsj csr
Other assigned terms:
- abbreviations
- acoustic model
- ambiguity
- association for computational linguistics
- benchmark
- benchmark corpora
- bias
- bigram
- case
- case information
- complete parse
- concept
- conditional model
- conditional probability
- constraint dependency grammar
- context-free grammar
- continuous speech
- corpora
- correlation
- correlations
- data set
- data sparseness
- data sparsity
- data structure
- dependency grammar
- dependency grammars
- dependency relations
- dependency structures
- dependency treebank
- development set
- distribution
- error rate
- estimation
- fact
- feature
- governor
- grammar
- grammar model
- grammar rule
- grammars
- grammatical features
- grammatical function
- grammaticality
- hypotheses
- hypothesis
- interpolation
- joint probability
- knowledge
- language model
- language model quality
- language models
- lattice
- lattices
- lemma
- lexical categories
- lexical category
- lexical feature
- lexical features
- lexical information
- lexicalized tree-adjoining grammar
- lexicon
- likelihood
- linguistic
- linguistic structure
- linguistics
- link direction information
- link type information
- method
- methodology
- mood
- morphological features
- n-gram
- n-grams
- nonterminals
- parse
- parse structure
- parse tree
- parsing accuracy
- partof-speech
- penn treebank
- perplexity
- perplexity reduction
- prague dependency treebank
- probabilistic model
- probabilities
- probability
- probability distributions
- probability model
- procedure
- production rules
- pronoun
- ptb
- punctuation
- recognition task
- relation
- semantic
- semantic constraints
- sentence
- sentences
- sparseness problem
- speech corpus
- speech data
- speech recognition problem
- speech recognition task
- structural information
- superarv
- superarv structure
- supertag
- symbol
- symbols
- syntactic constraints
- syntactic information
- syntactic knowledge
- syntax
- tag sequence
- tags
- technique
- test set
- text
- text corpus
- training
- training data
- training set
- transformation
- tree
- tree-adjoining grammar
- treebank
- trees
- trigram
- utterance
- valency
- vocabulary
- vocabulary size
- word
- word distribution
- word error rate
- word form
- word information
- word lattice
- word level
- word perplexity
- word sequence
- word sequences
- word usage
- words
