Vector Space Modelling and Latent Semantic Indexing for textual information retrieval
MUZAFAR RASOOL BHAT
Department of Computer Science, Islamic University of Science and Technology, Awantipora, Sriagar (India)
information retrieval, SVD, VSM, LSI
Vector space model of a collection eases process of information retrieval. This process can be further simplified by decomposing term document matrix using singular value decomposition. Use of appropriate number of singular values produced by singular value decomposition can improve the relevance of retrieved results. This paper compares the results produced by two main algebraic model based algorithms namely VSM and LSI using metrics based on Precision and Recall tests.Commonly used collections like CACM, and CISI is used in this research.
2455-9997 (Online) - 2229-3531 (Print)
Copy the following to cite this article: