ACL RD-TEC 1.0 Summarization of C04-1190
Paper Title:
SEMI-SUPERVISED TRAINING OF A KERNEL PCA-BASED MODEL FOR WORD SENSE DISAMBIGUATION
SEMI-SUPERVISED TRAINING OF A KERNEL PCA-BASED MODEL FOR WORD SENSE DISAMBIGUATION
Authors: Weifeng Su and Marine Carpuat and Dekai Wu
Primarily assigned technology terms:
- algorithm
- analysis technique
- classification
- classifiers
- dimensionality reduction
- disambiguation
- disambiguation method
- error analysis
- kernel
- kernel pca
- kpca
- kpca training
- learning
- maximum entropy
- maximum entropy model
- modeling
- modelling
- nearest neighbors
- nlp
- nonlinear principal component extraction
- polynomial kernel
- principal component analysis
- semi-supervised training
- sense disambiguation
- supervised classification
- supervised learning
- support vector machines
- unsupervised training
- vector representation
- word sense disambiguation
Other assigned terms:
- annotated training set
- approach
- baseline model
- bayes model
- case
- classification task
- component vector
- corpora
- data set
- data sets
- data sparseness
- dimensionality
- distribution
- entropy
- entropy models
- evaluations
- fact
- feature
- feature information
- feature set
- feature space
- highdimensional feature space
- hypothesis
- interpretation
- kernel function
- knowledge
- kpca model
- kpca-based model
- labeled training data
- learning model
- linear order
- mapping
- maps
- maximum entropy models
- method
- most-frequent-sense baseline
- natural language
- nlp tasks
- noise
- precision
- probability
- procedure
- relative frequency
- sense distinctions
- sense granularity
- senseval-2 data set
- sentence
- sentences
- sparse data
- stems
- supervised model
- support vector
- synonyms
- target word
- technique
- terms
- test data
- text
- theory
- tipster corpus
- training
- training corpora
- training corpus
- training data
- training instance
- training set
- unannotated corpora
- verb
- word
- word sense
- wordnet
- wordnet senses
- words
- wsd model
