ACL RD-TEC 1.0 Summarization of P96-1029
Paper Title:
COMPILATION OF WEIGHTED FINITE-STATE TRANSDUCERS FROM DECISION TREES
COMPILATION OF WEIGHTED FINITE-STATE TRANSDUCERS FROM DECISION TREES
Authors: Richard Sproat and Michael Riley
Primarily assigned technology terms:
- algorithm
- classification
- computing
- continuous speech recognition
- database
- decision tree
- decision trees
- finite-state transducers
- finitestate
- hidden markov
- hidden markov models
- kaplan-kay
- modeling
- nlp
- off-line compilation
- processing
- recognition
- regression
- regression trees
- regular expression
- resource management
- rewrite-rule rule-compilation
- rule-compilation
- speech processing
- speech recognition
- speech synthesis
- synthesis
- text-to-speech
- transducer
- transducers
- tree-based modeling
- weighted finite-state transducers
Other assigned terms:
- alphabet
- case
- composition
- computational framework
- continuous speech
- fact
- forest
- grammars
- input string
- interpretation
- interpreter
- language model
- lattice
- leaf
- lexicon
- linguistic
- mapping
- mappings
- markov models
- mechanisms
- method
- phoneme
- phonemes
- phonological rules
- probabilities
- probability
- pronunciation
- recognition model
- recognition task
- regular expressions
- rewrite rules
- root node
- runtime
- segments
- sentences
- speech recognition task
- stress
- symbol
- symbols
- terms
- timit database
- tokens
- training
- training data
- tree
- tree node
- trees
- two-level rule
- word
- word boundary
- words
