Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
ON THE ROAD TO IMPROVED LEXICAL CONFUSABILITY METRICS Eric Fosler-Lussier, Ingunn Amdal
 

Summary: ON THE ROAD TO IMPROVED LEXICAL CONFUSABILITY METRICS
Eric Fosler-Lussier, Ingunn Amdal£
, and Hong-Kwang Jeff Kuo
Bell Labs, Lucent Technologies
600 Mountain Ave.
Murray Hill, NJ 07974 USA
fosler,kuo@research.bell-labs.com
ABSTRACT
Pronunciation modeling in automatic speech recognition
systems has had mixed results in the past; one likely rea-
son for poor performance is the increased confusability in
the lexicon from adding new pronunciation variants. In this
work, we propose a new framework for determining lexi-
cally confusable words based on inverted finite state trans-
ducers (FSTs); we also present experiments designed to test
some of the implementation details of this framework. The
method is evaluated by looking at how well the algorithm
predicts the errors in an ASR system. We see from the con-
fusions learned in a training set that we are able to general-
ize this information to predict errors in an unseen test set.

  

Source: Amdal, Ingunn - Department of Electronics and Telecommunications, Norwegian University of Science and Technology

 

Collections: Engineering