-
NL Seminar- Greg Ver Steeg:
Fri, Nov 01, 2013 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Greg Ver Steeg, USC/ISI
Series: Natural Language Seminar
Abstract: Because natural language is complex, researchers in many domains look for lower-dimensional representations of text to suit their purposes. Different methods attempt to single out intuitive aspects of language like content, sentiment, or style. I will discuss a new, unsupervised approach to learning abstract representations of text (or other high-dimensional signals). The motivating principle is to use information theory to construct higher-order features that explain correlations between lower-order features. I will present preliminary results using this framework.
Biography: Greg Ver Steeg is a research professor at ISI. His research explores practical methods for inferring meaningful structure in complex systems like social networks. He did his PhD in quantum physics at Caltech.
Home Page:
http://www.isi.edu/people/gregv/about
Host: Yang Gao
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
Audiences: Everyone Is Invited
Contact: Peter Zamar
Event Link: http://nlg.isi.edu/nl-seminar/
