ACL RD-TEC 1.0 Summarization of W03-1020
Paper Title:
A FAST ALGORITHM FOR FEATURE SELECTION IN CONDITIONAL MAXIMUM ENTROPY MODELING
A FAST ALGORITHM FOR FEATURE SELECTION IN CONDITIONAL MAXIMUM ENTROPY MODELING
Authors: Yaqian Zhou and Fuliang Weng and Lide Wu and Hauke Schmidt
Primarily assigned technology terms:
- algorithm
- approximation
- binary search
- categorization
- chunking
- computing
- conditional maximum entropy
- cutoff
- cutoff method
- estimation process
- feature selection
- feature selection algorithm
- feature selection process
- greedy search
- ifs feature selection
- incremental feature selection
- language modeling
- language processing
- linear regression
- maximum entropy
- me modeling
- me training
- model selection
- modeling
- natural language processing
- normalization
- np chunking
- np recognition
- optimization
- parameter estimate
- parameter estimation
- processing
- recognition
- regression
- search
- selection algorithm
- selection process
- selection technique
- text categorization
- weighting
Other assigned terms:
- array
- case
- conditional probabilities
- data set
- distribution
- entropy
- estimation
- exponential model
- feature
- feature set
- feature sets
- feature space
- grammatical relation
- knowledge
- large feature space
- likelihood
- linguistic
- linguistic knowledge
- measure
- method
- natural language
- normalization factor
- optimization criterion
- ordered list
- paragraph
- parameter values
- part-of-speech
- part-of-speech tags
- penn treebank
- phrase
- phrase attachment
- pos tag
- precision
- prepositional phrase
- prepositional phrase attachment
- prior distribution
- probabilities
- probability
- process
- relation
- scalability
- search tree
- tag information
- tags
- technique
- technology
- termination condition
- test corpus
- test data
- text
- training
- training data
- training instance
- tree
- treebank
- understanding
- uniform distribution
- word
- words
