Data processing: speech signal processing – linguistics – language – Speech signal processing – Application
Reexamination Certificate
2000-09-09
2002-06-11
Ŝmits, T{overscore (a)}livaldis Ivars (Department: 2654)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Application
C704S254000, C707S793000
Reexamination Certificate
active
06405172
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates generally to voice speech recognition, and more particularly, but not exclusively, to the retrieval of records from a directory using spoken characters.
Certain modern data retrieval systems use voice speech recognition technology to select a desired record from among many. These systems, however, fail to perform adequately in certain circumstances, such as in the recognition of certain characters that sound similar when spoken. Such failures severely limit the utility of these systems for many operators and in many applications.
Other systems fail to correctly retrieve records when one or more characters are missing or incorrectly interpreted. Again, such systems are of limited utility in many applications and for many operators.
It is, therefore, apparent that a need exists for improved systems that apply voice speech recognition technology to large-directory look-up situations.
SUMMARY OF THE INVENTION
It is an object of this invention to provide an improved system for retrieving records from a directory using spoken characters as input.
It is another object of this invention to provide an improved table look-up system for contexts in which operators speech patterns are inconsistent, or the prefix letters that are read by the operator are not clearly legible.
These objects and others are provided in a system, method, and apparatus that retrieve data from a directory based on the spoken initial characters of one or more fields. Substitution groups are established, each containing characters that sound alike when spoken. For each query, an operator speaks the first few characters of the one or more fields. The characters are parsed and decoded from the speech, thereby producing a set of candidate decodings for each character position. Then, for at least one character position, one or more alternative characters (from the same substitution group(s) as the candidate character(s) for that character position) are selected to broaden the search. In some such embodiments, a regular expression is created that, for each character position output by the voice engine, matches (1) any of the candidate characters presented by the voice engine, or (2) any alternative character that is in a substitution group within one or more of the decoded characters. The regular expression is processed by an inexact string matching look-up routine and applied to the directory. The best matches are presented to the operator, who selects the desired record.
Other embodiments, forms, variations, objects, features, and applications may appear to those skilled in the art from the drawings and description contained herein.
REFERENCES:
patent: 4276597 (1981-06-01), Dissly et al.
patent: 4453217 (1984-06-01), Boivie
patent: 4556944 (1985-12-01), Daniels et al.
patent: 4632252 (1986-12-01), Haruki et al.
patent: 4866778 (1989-09-01), Baker
patent: 4908864 (1990-03-01), Togawa et al.
patent: 4921107 (1990-05-01), Hofer
patent: 4979206 (1990-12-01), Padden et al.
patent: 5101375 (1992-03-01), Goldhor
patent: 5212730 (1993-05-01), Wheatley et al.
patent: 5263118 (1993-11-01), Cornelison
patent: 5353221 (1994-10-01), Kutsumi et al.
patent: 5454063 (1995-09-01), Rossides
patent: 5500920 (1996-03-01), Kupiec
patent: 5581599 (1996-12-01), Tsuji et al.
patent: 5677834 (1997-10-01), Mooneyham
patent: 5677990 (1997-10-01), Junqua
patent: 5752195 (1998-05-01), Tsuji et al.
patent: 5752230 (1998-05-01), Alonso-Cedo
patent: 5905773 (1999-05-01), Wong
patent: 5987410 (1999-11-01), Kellner et al.
patent: 5995928 (1999-11-01), Nguyen et al.
patent: 6009392 (1999-12-01), Kanevsky et al.
patent: 6032164 (2000-02-01), Tsai
patent: 6052439 (2000-04-01), Gerszberg et al.
patent: 6317489 (2001-11-01), Parsadayan
patent: 876036 (1998-11-01), None
patent: 99/09729 (1999-02-01), None
Auberry Eric R.
Baker Christopher A.
Baker Peter N.
Moon Alexander R.
Starrett Cortland D.
Ŝmits T{overscore (a)}livaldis Ivars
Chaclas Angelo N.
MailCode Inc.
Malandra Charles R.
Vitale Alberta A.
LandOfFree
Voice-enabled directory look-up based on recognized spoken... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Voice-enabled directory look-up based on recognized spoken..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Voice-enabled directory look-up based on recognized spoken... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2973773