An accurate model for Latent Semantic Analysis (LSA) by Singular Value Decomposition (SVD)

An accurate model for Latent Semantic Analysis (LSA) by Singular Value Decomposition (SVD)

Neda Mohammadi1 Hadi Mehdipour2

1) Department of Computer Engineering and IT, Shiraz University of Technology, Shiraz, Iran Email:
2) Department of Shahid Chamran, Technical and Vocational University, Kerman, Iran Email:

IOI Code : XYCC-ABKKB
Publication : International Conference on Science and Engineering(icesconf.com)
Abstract :
    This paper presents a conceptual model to retrieve latent semantic automatically. The model specifies relations among documents and query based on latent semantic mapping. To achieve the goal, Singular Value Decomposition (SVD) and the concept of eigenvectors and eigenvalues is used which they are a theory in linear algebra. SVD has many usage in science and engineering. Many efforts is done to retrieve latent semantic and relation among texts, images and etc. By singular value decomposition can extract latent semantic between text or images with higher accuracy. This paper uses SVD to extract the latent semantic between documents and queries. Results show which SVD provides a complete automatic method to retrieve latent semantics.

Keywords : singular value decomposition (svd) latent semantic analysis (lsa) eigenvalues eigenvectors