LSI vs. Wordnet Ontology in Dimension Reduction for Information Retrieval.
Pavel MoravecMichal KolovratVáclav SnáselPublished in: DATESO (2004)
Keyphrases
- dimension reduction
- wordnet
- latent semantic indexing
- semantic information
- information retrieval
- singular value decomposition
- vector space model
- lexical database
- lexical resources
- semantic relationships
- domain specific ontologies
- knowledge base
- semantic similarity
- semantic relations
- natural language processing
- co occurrence
- word sense disambiguation
- ontology learning
- feature extraction
- high dimensional
- text retrieval
- semantic relatedness
- domain ontology
- principal component analysis
- computational linguistics
- low dimensional
- text documents
- relatedness measure
- preprocessing
- high dimensional data
- knowledge representation
- linear discriminant analysis
- vector space
- search engine
- document representation
- least squares
- information retrieval systems
- dimensionality reduction
- text mining
- semantic web
- high dimensionality
- cluster analysis
- retrieval systems
- test collection
- unsupervised learning
- feature selection
- data points
- keywords
- document collections
- data sets
- machine learning
- learning algorithm
- text classification
- feature space
- pairwise
- tf idf
- language model
- information extraction
- relevance feedback
- nearest neighbor
- document retrieval