Solving total least-squares problems in information retrieval

DOI10.1016/S0024-3795(00)00030-6MaRDI QIDQ1587280zbMATH OpenFDO

Authors Eric P. Jiang, Michael W. Berry

Publication date 17 May 2001

Published in Linear Algebra and its Applications (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/S0024-3795(00)00030-6

zbMATH Keywords

algorithminformation filteringinformation retrievallatent semantic indexingnumerical experimentsRiemannian singular value decomposition

Mathematics Subject Classification ID

Numerical solutions to overdetermined systems, pseudoinverses (65F20) Information storage and retrieval of data (68P20)

Review text

The Riemannian singular value decomposition (R-SVD) is modified and used to formulate an enhanced implementation of latent semantic indexing (LSI) for conceptual information retrieval. A new algorithm for computing the R-SVD is also described. In updating the LSI models, this R-SVD can be very effective. Experiments demonstrate that a 20\% improvement (in retrieval) over the current LSI model is possible.

Recommendations

• scientific article; zbMATH DE number 1748566
• Lanczos and the Riemannian SVD in information retrieval applications
• scientific article; zbMATH DE number 2069285
• On Updating Problems in Latent Semantic Indexing
• Two uses for updating the partial singular value decomposition in latent semantic indexing

Cites work

• scientific article; zbMATH DE number 51511 (Why is no real title available?)
• scientific article; zbMATH DE number 961607 (Why is no real title available?)
• Structured total least squares and \(L_ 2\) approximation problems
• SYMMETRIC GAUGE FUNCTIONS AND UNITARILY INVARIANT NORMS
• The Restricted Singular Value Decomposition of Matrix Triplets
• Total Least Norm Formulation and Solution for Structured Problems
• Using Linear Algebra for Intelligent Information Retrieval

Cited in

(6)

• Lightweight relevance filtering for machine-generated resolution problems
• Lanczos and the Riemannian SVD in information retrieval applications
• scientific article; zbMATH DE number 2069285 (Why is no real title available?)
• scientific article; zbMATH DE number 1748566 (Why is no real title available?)
• An ensemble Kalman filter implementation based on modified Cholesky decomposition for inverse covariance matrix estimation
• scientific article; zbMATH DE number 782060 (Why is no real title available?)

Describes a project that uses

Uses Software

• VanHuffel
• svdpack
• Okapi

This page was built for publication: Solving total least-squares problems in information retrieval

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1587280)