[Colloquium] Today: Matveeva/Dissertation Defense/2-19-08
Margaret Jaffey
margaret at cs.uchicago.edu
Tue Feb 19 10:09:44 CST 2008
A reminder about Irina's defense this afternoon.
Department of Computer Science/The University of Chicago
*** Dissertation Defense ***
Candidate: Irina Matveeva
Date: Tuesday, February 19, 2008
Time and Location: 1:00 p.m. in Ryerson 255
Title: Generalized Latent Semantic Analysis for Document Representation
Abstract:
Document representation has a large impact on the performance of
document retrieval and clustering algorithms. Since content words can
be combined into semantic classes there has been a considerable
interest in low-dimensional term and document representations. We
developed the Generalized Latent Semantic Analysis (GLSA) framework to
extend the applicability of the idea of the latent semantic space
analysis with spectral embedding. GLSA uses linguistically motivated
measures of similarity between words and employs different
dimensionality reduction techniques to preserve them.
Most language modelling and dimensionality reduction approaches are
applied to the full vocabulary. However, semantics and syntax provide
a theoretical justification for treating subsets of the vocabulary
differently. Moreover, the theoretical justification of GLSA suggest
that it is most appropriate applied to a subset of the vocabulary
comprised of content words. Therefore, I propose to use GLSA with a
hybrid approach to document indexing. My goal is to keep the notion of
latent semantic concepts and at the same time preserve the specifics
of the document collection. Our experiments show that GLSA improves
the results of the similarity based document classification and
clustering algorithms and text segmentation algorithms.
Candidate's Advisers: Gina-Anne Levow, Partha Niyogi
A draft copy of Ms. Matveeva's dissertation is available in Ry 156.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Margaret P. Jaffey margaret at cs.uchicago.edu
Department of Computer Science
Student Support Rep (Ry 156) (773) 702-6011
The University of Chicago http://www.cs.uchicago.edu
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
More information about the Colloquium
mailing list