CLSI: A Flexible Approximation Scheme from Clustered Term-Document Matrices.
Efstratios GallopoulosDimitrios ZeimpekisPublished in: SDM (2005)
Keyphrases
- polynomial time approximation
- document images
- approximation schemes
- document representation
- inverted lists
- information retrieval
- indexing scheme
- approximation algorithms
- term frequency
- approximation error
- index terms
- query terms
- singular value decomposition
- web documents
- matrix representation
- structured documents
- keywords
- original data
- semantic information
- document collections
- vector space model
- document classification
- relevance model
- document clustering
- initial query
- text representation
- text documents
- retrieval systems
- document space
- polynomial approximation
- posterior marginals
- randomized approximation