Showing papers on "Latent semantic analysis published in 1995"
TL;DR: A lexical match between words in users’ requests and those in or assigned to documents in a database helps retrieve textual materials from scientific databases.
Abstract: Currently, most approaches to retrieving textual materials from scientific databases depend on a lexical match between words in users’ requests and those in or assigned to documents in a database. ...
1,630 citations
TL;DR: The focus of this work is to demonstrate the computational advantages of exploiting low-rank orthogonal decompositions such as the ULV (or URV) as opposed to the truncated singular value decomposition (SVD) for the construction of initial and updated rank-k subspaces arising from LSI applications.
Abstract: Current methods to index and retrieve documents from databases usually depend on a lexical match between query terms and keywords extracted from documents in a database. These methods can produce incomplete or irrelevant results due to the use of synonyms and polysemus words. The association of terms with documents (or implicit semantic structure) can be derived using large sparse {\it term-by-document} matrices. In fact, both terms and documents can be matched with user queries using representations in k-space (where 100 ≤ k ≤ 200) derived from k of the largest approximate singular vectors of these term-by-document matrices. This completely automated approach called latent semantic indexing or LSI, uses subspaces spanned by the approximate singular vectors to encode important associative relationships between terms and documents in k-space. Using LSI, two or more documents may be closeto each other in k-space (and hence meaning) yet share no common terms. The focus of this work is to demonstrate the computational advantages of exploiting low-rank orthogonal decompositions such as the ULV (or URV) as opposed to the truncated singular value decomposition (SVD) for the construction of initial and updated rank-k subspaces arising from LSI applications.
79 citations
01 Jan 1995
TL;DR: Based on data collected from the usage of the system by graduate students and University of Tennessee library patrons, LSIRS is shown to be an effective and useful document retrieval system for both the inexperienced and advanced user.
Abstract: In this report, a study and analysis of the e ectiveness of the Latent Semantic Indexing Retrieval System (LSIRS) is presented. Using a Motif-based X-Windows application, LSIRS uses the truncated singular value decomposition (SVD) of the associated term-document matrices to perform document retrieval. The LSIRS user interface was initial prototype graphical user interface (GUI) of the recently developed XLSI application. The indexing and SVD software used to employ Latent Semantic Indexing (or LSI) was developed at Bellcore and the University of Tennessee. Based on data collected from the usage of the system by graduate students and University of Tennessee library patrons, LSIRS is shown to be an e ective and useful document retrieval system for both the inexperienced and advanced user. Suggestions for future system improvements are also described. i
11 citations