ACL RD-TEC 1.0 Summarization of P00-1014
Paper Title:
AN UNSUPERVISED APPROACH TO PREPOSITIONAL PHRASE ATTACHMENT USING CONTEXTUALLY SIMILAR WORDS
AN UNSUPERVISED APPROACH TO PREPOSITIONAL PHRASE ATTACHMENT USING CONTEXTUALLY SIMILAR WORDS
Authors: Patrick Pantel and Dekang Lin
Primarily assigned technology terms:
- algorithm
- approximation
- back-off estimation
- bayes estimation
- classification
- classifier
- classifiers
- corpus-based approach
- data extraction
- database
- disambiguation
- internet
- iterative algorithm
- iterative process
- language disambiguation
- language learning
- language processing
- learning
- learning algorithms
- minipar
- naive bayes
- natural language disambiguation
- natural language processing
- parser
- phrase attachment algorithm
- processing
- sense disambiguation
- supervised learning
- transformation-based learning
- unsupervised approach
- unsupervised method
- word sense disambiguation
Other assigned terms:
- ambiguity
- approach
- attachment score
- attachment site
- beam
- break
- classification model
- classification task
- clusters
- co-occurrence
- co-occurrence frequency
- collocation
- conditional probabilities
- data flow
- data set
- data sets
- dependency relationship
- dependency tree
- dependency trees
- estimation
- events
- experimental results
- feature
- feature space
- head noun
- heuristic
- large corpus
- lexical item
- lexicon
- linear combination
- meaning
- method
- modifier
- mutual information
- natural language
- newspaper corpus
- noun phrase
- nouns
- parsed corpus
- penn treebank
- performance comparison
- phrase
- phrase attachment
- precision
- preposition
- prepositional attachment
- prepositional phrase
- prepositional phrase attachment
- prepositions
- probabilities
- process
- russian
- sense inventory
- sentence
- sentences
- sparse data
- test data
- test set
- testing corpus
- thesaurus
- training
- training data
- training data set
- training set
- tree
- treebank
- trees
- verb
- verb phrase
- word
- word sense
- words
