LSA and Information Retrieval

Getting Back to Basics

Authored by: Susan T. Dumais

Handbook of Latent Semantic Analysis

Print publication date:  February  2007
Online publication date:  May  2013

Print ISBN: 9780805854183
eBook ISBN: 9780203936399
Adobe ISBN: 9781135603281


 Download Chapter



Latent semantic analysis (LSA) was first introduced in Dumais, Furnas, Landauer, Deerwester, and Harshman (1988) and Deerwester, Dumais, Furnas, Landauer, and Harshman (1990) as a technique for improving information retrieval. Most web, enterprise, and desktop search engines work by matching words in a user ‘s query with words in documents. Information retrieval systems that depend on lexical matching suffer from two problems. First, because the same literal word can have many different meanings, irrelevant information is retrieved when searching. Second, because there are different ways to describe the same object or concept, relevant information can be missed. LSA was designed to address these fundamental retrieval problems. The key idea in LSA was to use dimension reduction techniques to map documents and terms into a lower dimensional semantic space.

Search for more...
Back to top

Use of cookies on this website

We are using cookies to provide statistics that help us give you the best experience of our site. You can find out more in our Privacy Policy. By continuing to use the site you are agreeing to our use of cookies.