AI Seminar
October 14, 2002
Melanie Martin

Introduction to Latent Semantic Analysis

Abstract:

Latent Semantic Analysis was developed at Bellcore in the late 1980s, as
an improvement on vector space methods used for Information Retrieval.
Latent Semantic Analysis has proved useful many areas of Natural Language
Processing, Information Retrieval and Psychology.

In this talk we will discuss the motivation and methodology for Latent
Semantic Analysis and lay the foundation for the "Applications of Latent
Semantic Analysis" talk by Peter Foltz on October 28th.