Ex Parte Gillick et al

Patent Trial and Appeal Board

Aug 31, 2017

13338383 (P.T.A.B. Aug. 31, 2017)

United States Patent and Trademark Office
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O.Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
13/338,383 12/28/2011 Laurence Gillick 30054-003001 5872
69713 7590 09/05/2017
OCCHIUTI & ROHLICEK LLP
321 Summer St.
Boston, MA 02210
EXAMINER
HONG, THOMAS J
ART UNIT PAPER NUMBER
3715
NOTIFICATION DATE DELIVERY MODE
09/05/2017 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
INFO @ORP ATENT.COM
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte LAURENCE GILLICK, ALAN SCHWARTZ,
JEAN-MANUEL VAN THONG, PETER WOLF, and
DON MCALLASTER
Appeal 2015-001023
Application 13/338,383
Technology Center 3700
Before JOHN C. KERINS, GEORGE R. HOSKINS, and
BRANDON J. WARNER, Administrative Patent Judges.
KERINS, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Laurence Gillick et al. (Appellants) appeal under 35 U.S.C. § 134(a)
from a final rejection of claims 1—14. We have jurisdiction under 35 U.S.C.
§ 6(b).
We REVERSE.
Appeal 2015-001023
Application 13/338,383
THE INVENTION
Appellants’ invention is directed to a method for automated
processing of speech in a speech training system. Claim 1, reproduced
below, is illustrative:
1. A method for automated processing of a user’s speech in
a speech training system, the method comprising:
accepting a data representation of a user’s speech; and
processing the data representation of the user’s speech
according to a statistical model, said model comprising model
parameters associated with each of a plurality of speech units,
the model parameters associated with at least some of the
speech units including parameters associated with target
instances of the speech unit and parameters associated with
non-target instances of the speech unit;
wherein the processing includes determining quantities
characterizing degrees of match of the data to speech units
represented in the statistical model and determining an
aggregated measure of one or more classes of speech errors in
the user’s speech from the determined quantities.
REJECTIONS
The Examiner rejects:
(i) claims 1, 3—6, 11, 13, and 141 under 35 U.S.C. § 103(a) as being
unpatentable over Faisman (US 2006/0161434 Al, published July 20, 2006)
in view of Yu (US 2009/0258333 Al, published Oct. 15, 2009); and
1 Claim 14 is not listed in the initial statement of the rejection, but is
addressed in the body of the rejection. We regard the initial omission as a
harmless error.
2
Appeal 2015-001023
Application 13/338,383
(ii) claims 2, 7—10, and 12 under 35 U.S.C. § 103(a) as being
unpatentable over Faisman in view of Yu and Johnson (US 2009/0004633
Al, published Jan. 1, 2009).
OPINION
Claims 1, 3—6, 11, 13, and 14—Unpatentability over Faisman and Yu
The Examiner states, with respect to independent claims 1,11, and 13,
that “Faisman discloses a method/system/tangible machine readable medium
for automated processing of a user’s speech in a speech training system.”
Final Act. 3. The Examiner acknowledges that Faisman does not disclose
that the processing includes “determining quantities characterizing degrees
of match of the data to speech units represented in the statistical model and
determining an aggregated measure of one or more classes of speech errors
in the user’s speech from the determined quantities.” Id.
The Examiner turns to Yu as teaching a statistical model to analyze
the learner’s input, and as teaching that the likelihood of matching between
the input and the model speech is quantified. Final Act. 3—4, citing Yu, para.
47,11. 11—12, para. 49,11. 6—11. The Examiner additionally finds that
paragraph 10 of Yu “teaches how the calculated quantities are converted to
scores for determination.” Id. at 4. The Examiner then concludes that it
would have been obvious to modify Faisman, presumably to incorporate
these features from Yu, “to improve techniques related to machine based
spoken language learning.” Id., citing Yu, paras. 2, 4.
Appellants argue, inter alia, that the Examiner “has not identified how
specifically any degree of match to text units ... is taught in Yu,” such that
the teachings could be reasonably incorporated into Faisman. Br. 6.
Appellants also assert that the Examiner “has failed to identify any specific
3
Appeal 2015-001023
Application 13/338,383
‘aggregate measure’ in Yu that is based on Yu’s purported ‘degree of match’
quantities for text units.” Id. The Examiner does not respond to these
specific points in the Answer. The Examiner does identify, in an Advisory
Action dated November 27, 2013, additional portions of the Yu reference
where certain claim terms and/or terms used in Appellants’ arguments can
be found.
The Examiner fails to explain adequately how Faisman is to be
modified in view of the teachings of Yu, or how such modifications will
improve, in Faisman, “techniques related to machine based spoken language
learning,” the reason articulated by the Examiner for making such
modifications. Final Act. 4. At the outset, we note that the underlying
premise of the Examiner’s rejection, i.e., that Faisman is directed to a speech
training system, is flawed. The Examiner points to the Abstract of Faisman
as evidencing this to be the purpose of the Faisman system, however, the
Abstract simply refers to the provision of a method for improving spoken
language. The remainder of the Faisman disclosure evidences that the
improvement sought is not that of training the user to improve his or her
spoken language capabilities, but rather a correction of detected errors in a
person’s speech (converted to and processed as text data) to produce an
improved text and/or speech output to other readers/listeners. Faisman,
paras. 19, 20. As such, although possible, it is not readily apparent how the
proposed modification(s) are even applicable to the operation of the Faisman
system, let alone how they would improve the operation of Faisman. The
discussion of Yu presented in the rejection further appears to simply point to
passages in Yu that incidentally use the same or similar terminology as do
4
Appeal 2015-001023
Application 13/338,383
Appellants’ claims, without adequate analysis as to how these aspects of the
Yu system would improve the Faisman system, as asserted in the rejection.
Independent claim 14 includes a processing step of determining a
confidence measure of the speaker exhibiting a class of speech errors, and
determining if the confidence measure exceeds a threshold. Final Act. 5—6.
The Examiner cites to passages in Yu that are said to teach these limitations,
but again, the Examiner has not adequately explained the possible relevance
of these features to the Faisman system. The articulated reason for making
the modification, namely to improve techniques related to machine based
spoken language learning, does not have a clear nexus to the Faisman
system, which is not a language training system.
Accordingly, we do not sustain the rejection of claims 1, 3—6, 11, 13,
and 14 as being unpatentable over Faisman and Yu.
Claims 2, 7—10, and 12—Unpatentability over Faisman, Yu, and Johnson
The Examiner does not rely on Johnson in any manner that would
appear to overcome the deficiencies noted above with respect to the
proposed combination of Faisman and Yu. We do not sustain the rejection
of claims 2, 7—10, and 12 as being unpatentable over Faisman, Yu, and
Johnson.
DECISION
The rejections of claims 1—14 under 35 U.S.C. § 103(a) are reversed.
REVERSED
5