ACL RD-TEC 1.0 Summarization of P02-1001
Paper Title:
PARAMETER ESTIMATION FOR PROBABILISTIC FINITE-STATE TRANSDUCERS
PARAMETER ESTIMATION FOR PROBABILISTIC FINITE-STATE TRANSDUCERS
Primarily assigned technology terms:
- abstracting
- algorithm
- computing
- conditional random fields
- discriminative training
- em algorithm
- estimation method
- feature selection
- final state
- finite-state automata
- finite-state framework
- finite-state transducers
- forward-backward algorithm
- gradient-based optimization
- hardware
- hmms
- iterative scaling
- maximum-entropy
- maximum-likelihood
- maximum-likelihood estimation
- model merging
- model training
- modeling
- noisy-channel model
- normalization
- optimization
- parameter estimation
- parameter tying
- parameterization
- partitioning
- probabilistic finite-state
- pruning
- random walk
- random-walk
- regular expression
- supervised training
- training algorithm
- transcription
- transducers
- viterbi
- weighted finite-state transducers
Other assigned terms:
- acyclic graph
- alphabet
- approach
- automata
- bias
- case
- composition
- conditional distribution
- data structure
- device
- distribution
- edit distance
- estimation
- expectation semiring
- fact
- feature
- implementation
- input string
- interpretation
- joint distribution
- linguistic
- local maxima
- log-linear model
- mapping
- maps
- method
- noisy channel
- parallelism
- pcfg
- pcfgs
- posterior
- posterior probability
- prior probability
- probabilities
- probability
- probability distribution
- probability distributions
- recipe
- regular expressions
- relation
- relaxation technique
- semiring
- statistical models
- subgraph
- suffixes
- symbol
- symbols
- technique
- terms
- theorem
- topology
- training
- training data
- transcriptions
- tree
- user
- vector space
