Browse Prior Art Database

Method for Finding and Correcting Homonyms in Dictated Text

IP.com Disclosure Number: IPCOM000123605D
Original Publication Date: 1999-Feb-01
Included in the Prior Art Database: 2005-Apr-05
Document File: 1 page(s) / 42K

Publishing Venue

IBM

Related People

Ballard, B: AUTHOR [+3]

Abstract

Disclosed is method facilitating the correction of text, previously dictated into a speech recognition system, by highlighting words having homonyms in the text and by displaying alternative homonyms of the highlighted words.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Method for Finding and Correcting Homonyms in Dictated Text

   Disclosed is method facilitating the correction of
text, previously dictated into a speech recognition system, by
highlighting words having homonyms in the text and by displaying
alternative homonyms of the highlighted words.

   The figure shows a dialog box displayed when this feature
is selected.  A portion of the previously dictated text is shown in
the "Dictated text" list box, with any word having homonyms being
highlighted.  For each highlighted word, the homonyms are shown in
the "Alternative words" list box, in the form of a numbered list.

   The user can select one of these alternatives with his
pointing device, or he can make a selection in the "Dictated text"
list box, followed by selecting the "Change" button and typing his
correction.  If the user determines that the highlighted word shown
in the "Dictated text" list box is the correct choice, he selects the
"Skip" button.  In any case, the system then proceeds to the next
word, if any, having homonyms, so that the process can be repeated.

   This method is applied to words which sound similar, such
as "pear" and "per" in the example of the figure, as well as to
words which are true homonyms.  A simple version of this method
displays homonyms and similar-sounding words found in a database; an
advanced version suppresses the display of alternatives when the
context of the dictated text indicates their use would be
unreasonable.