Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Automatic Determination of Phonetic Markov Word Models

IP.com Disclosure Number: IPCOM000100118D
Original Publication Date: 1990-Mar-01
Included in the Prior Art Database: 2005-Mar-15
Document File: 2 page(s) / 99K

Publishing Venue

IBM

Related People

Bahl, LR: AUTHOR [+4]

Abstract

A technique is described whereby speech recognition devices can automatically determine phonetic Markov word models. When a device is given the spelling and one or more utterances of a previously unseen word, the concept determines a phonetic Markov word model for the word.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Automatic Determination of Phonetic Markov Word Models

       A technique is described whereby speech recognition
devices can automatically determine phonetic Markov word models. When
a device is given the spelling and one or more utterances of a
previously unseen word, the concept determines a phonetic Markov word
model for the word.

      In speech recognition systems which employ Markov word models,
two types of word models (phonetic and fenemic) are prominent.
Phonetic word models, also known as phonetic baseforms, are usually
obtained from phonetic transcriptions as listed in a dictionary or as
provided by a phonetician. Fenemic word models, also known as fenemic
baseforms, can be obtained automatically from multiple utterances.

      Generally, when users personalize the recognition vocabulary by
adding their own words, it is desirable to obtain both fenemic and
phonetic baseforms for the added words.  While the fenemic baseforms
can be determined readily from one or more utterances of the added
words, the phonetic baseforms cannot.

      Although attempts to determine phonetic baseforms automatically
from one or more utterances have been attempted previously, the
success has been limited.  The methods have attempted to create
phonetic baseforms solely from the acoustic evidence provided by the
utterances. However, an additional source of information is provided
by the spelling of a word, which, if used in conjunction with the
acoustic information offers the potential of a significant
improvement in the quality of automatic phonetic baseforms.

      The concept described herein utilizes the given spelling of a
word and one or more sample utterances to provide a means of
determining the most probable phonetic baseforms of a word.

      Attempts to determine phonetic baseforms in this way have been
previously made (1).  However, the concept described herein employs
techniques to improve on the previous methods, such as through the
use of superior spelling-to-phone rules in the form of pylonic idiot-
systems and superior acoustic analysis through the use of multiple
utterances.

      First, the existence of spelling-to-sound rules is assumed in
the form of an idiot-system, that several utterances of the subject
word are available and that a fenemic baseform has been constructed.
The most probable phonetic baseform may be determined by decoding the
fenemic baseform wi...