ACL RD-TEC 1.0 Summarization of J96-1002
Paper Title:
A MAXIMUM ENTROPY APPROACH TO NATURAL LANGUAGE PROCESSING
A MAXIMUM ENTROPY APPROACH TO NATURAL LANGUAGE PROCESSING
Authors: Adam L. Bergert and Vincent J. Della Pietra and Stephen A. Della Pietra
Primarily assigned technology terms:
- algorithm
- approximation
- automatic feature selection
- automatic machine translation
- automatic method
- bilingual sense disambiguation
- clustering
- computational linguistics
- computer science
- computing
- constrained optimization
- context-sensitive modeling
- cross-validation
- decision tree
- decision-making
- disambiguation
- dynamic programming
- dynamic programming algorithm
- em algorithm
- feature selection
- feature selection algorithm
- french-to-english machine translation
- french-to-english translation
- incremental feature selection
- iterative scaling
- lagrange multiplier
- language modeling
- language processing
- likelihood training
- machine translation
- machine translation system
- maximum entropy
- maximum entropy approach
- maximum entropy method
- maximum entropy model
- maximum likelihood
- maximum likelihood training
- maximum-likelihood
- model selection
- modeling
- mutual-information
- natural language processing
- numerical optimization
- optimization
- optimization method
- parsing
- part-of-speech tagger
- pattern recognition
- preprocessing
- processing
- programming algorithm
- recognition
- recognition systems
- search
- searching
- segmentation
- segmenter
- selection algorithm
- sense disambiguation
- sentence segmentation
- speech recognition
- speech recognition systems
- spelling
- splitting
- statistical estimation
- statistical modeling
- statistical translation
- stochastic part-of-speech tagger
- surveillance
- tagger
- training process
- transcription
- translation modeling
- translation process
- translation system
- translator
- unconstrained optimization
- viterbi
- viterbi alignment
- word reordering
- word translation
- word-reordering
Other assigned terms:
- acoustic models
- approach
- association for computational linguistics
- bilingual corpus
- case
- computational complexity
- concept
- conditional distribution
- conditional probability
- conditional probability distribution
- context information
- contextual information
- convergence
- correlations
- data set
- decision-making process
- discourse
- distortion probability
- distribution
- empirical justification
- english language
- english sentence
- english translation
- english translations
- entropy
- entropy models
- estimation
- events
- expert knowledge
- exponential model
- fact
- feature
- finite set
- french
- french sentence
- french translation
- french word
- grammar
- grammars
- graphical representation
- hansard corpus
- hypothesis
- interpretation
- joint distribution
- joint probability
- knowledge
- language model
- likelihood
- linguistics
- log-likelihood
- mathematics
- maximum entropy models
- maximum entropy principle
- measure
- method
- modeling problem
- natural language
- noise
- noun phrase
- noun phrases
- nouns
- optimization problem
- parameter values
- parametric model
- part of speech
- part-of-speech
- part-of-speech tags
- passage
- phrase
- priori
- probabilities
- probability
- probability distribution
- probability distributions
- procedure
- process
- processing time
- pronunciation
- recipe
- relation
- reordering
- search space
- search strategy
- segment boundaries
- segment boundary
- segments
- sentence
- sentence pair
- sentences
- statistic
- statistical models
- statistics
- stochastic model
- symbol
- symbols
- tags
- technique
- technologies
- term
- terms
- test data
- text
- theorem
- theory
- training
- training data
- training sample ~
- training samples
- translation model
- translation models
- translation probabilities
- translation problem
- translations
- tree
- uniform distribution
- verb
- word
- word order
- word pronunciation
- words