ACL RD-TEC 1.0 Summarization of N06-1009
Paper Title:
ROLE OF LOCAL CONTEXT IN AUTOMATIC DEIDENTIFICATION OF UNGRAMMATICAL, FRAGMENTED TEXT
ROLE OF LOCAL CONTEXT IN AUTOMATIC DEIDENTIFICATION OF UNGRAMMATICAL, FRAGMENTED TEXT
Authors: Tawanda Sibanda and Ozlem Uzuner and Ozlem Uzuner
Primarily assigned technology terms:
- algorithm
- capitalization
- classification
- classifier
- classifiers
- cross-validation
- crossvalidation
- entity recognition
- hidden markov
- hidden markov models
- hmms
- identification
- kernel
- linear kernel
- linguistic processing
- maximum entropy
- maximum entropy model
- named entity recognition
- pattern-matching
- processing
- recognition
- rule-based approach
- rule-based system
- statistical approaches
- support vector machine
- support vector machines
- svm approach
- svm-based system
- ten-fold cross-validation
Other assigned terms:
- ambiguous word
- ambiguous words
- annotators
- approach
- bigram
- bigram language model
- biology
- biomedical domain
- case
- context information
- corpora
- data set
- device
- dictionaries
- dictionary
- distribution
- entity class
- entity type
- entity types
- entropy
- experimental results
- f-measure
- fact
- feature
- feature set
- feature sets
- foreign words
- head noun
- heuristic
- heuristic rules
- heuristics
- human annotators
- hypotheses
- hypothesis
- implementation
- information gain
- language model
- likelihood
- linguistic
- local context
- markov models
- mesh
- named entities
- named entity
- names
- news corpus
- part-of-speech
- part-of-speech tags
- precision
- probabilities
- probability
- proper names
- punctuation
- sentence
- sentences
- support vector
- svm implementation
- svms
- syntactic features
- tags
- target word
- technologies
- terms
- test corpus
- text
- tokens
- transition probabilities
- understanding
- unified medical language
- winnows
- word
- word formation
- words
