[comp.ai] Formalizing Knowledge used in Spectrogram Reading...

finin@prc.unisys.com (Tim Finin) (03/03/90)

			      AI SEMINAR
	  UNISYS CENTER FOR ADVANCED INFORMATION TECHNOLOGY
	       (formerly Unisys Paoli Research Center)

	  Formalizing Knowledge used in Spectrogram Reading:
	     Acoustic and perceptual evidence from stops

			    Lori F. Lamel
			      IBM France

An evaluation of spectrogram reading was conducted to obtain a better
understanding of the process of reading spectograms, to evaluate the
ability of spectrogram readers.  This talk has three parts: an
evaluation of spectrogram readers on a constrained task, a comparison
to listeners on the same task, and a formalization of
spectrogram-reading knowledge in a rule-based system.

A knowledge-based system for identifying stop consonants was
developed.  The framework was chosen because the expression and use of
the knowledge is explicit.  The emphasis was on capturing the acoustic
descriptions and modeling the reasoning thought to be used by human
spectrogram readers.  However, the implemention was much harder than
had been anticipated due to a variety of reasons.  The most important
is that there appears to be much more happening in our visual system
and in our thought processes than we actually express, even when asked
to explain our reasoning.  Humans are able to selectively pay
attention to acoustic evidence, even in the presence of contradictory
evidence.  This ability is not well understood, and is difficult to
mimic.

		    11:00 am Monday, March 5, 1990
			 CAIT Conference Room
	  Unisys Center for Advanced Information Technology
		     Great Valley Laboratories #1
			70 E. Swedesford Road
			    Paoli PA 19301
				   
   -- non-Unisys visitors who are interested in attending should --
   --   send email to finin@prc.unisys.com or call 215-648-2480  --
-- 
 Tim Finin                                   finin@prc.unisys.com
 Center for Advanced Information Technology  215-648-2840, 215-648-2288 (fax)
 Unisys, PO Box 517, Paoli, PA 19301         215-386-1749 (home)