Ex Parte Bennett et al

Patent Trial and Appeal Board

Dec 21, 2012

11030919 (P.T.A.B. Dec. 21, 2012)

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.
11/030,919 01/07/2005 Ian M. Bennett PHO99001C3 3099
7590 12/24/2012
Ian M. Bennett
634 Georgia Ave.
Palo Alto, CA 94306
EXAMINER
LERNER, MARTIN
ART UNIT PAPER NUMBER
2657
MAIL DATE DELIVERY MODE
12/24/2012 PAPER
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.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte IAN M. BENNETT, BANDI RAMESH BABU,
KISHOR MORKHANDIKAR, and PALLAKI GURURAJ
____________

Appeal 2009-013176
Application 11/030,9191
Technology Center 2600
____________

Before MARC S. HOFF, JAMES R. HUGHES, and
GREGORY J. GONSALVES, Administrative Patent Judges.

HOFF, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 2-15, 21-33, 35, and 36.2 We have jurisdiction under 35 U.S.C.
§ 6(b).
We affirm.

1 The real party in interest is Phoenix Solutions, Inc.
2 Claims 1, 20, and 34 have been cancelled. Claims 16-19 and 37-48 stand
withdrawn from consideration based on a restriction requirement.
Appeal 2009-013176
Application 11/030,919

2
STATEMENT OF THE CASE
Appellants’ invention is a system and method for responding to
speech based user inputs and queries presented over a distributed network
such as the Internet. A user can ask a question in a natural language such as
English, French, German, Spanish, or Japanese and receive the appropriate
answer at his or her computer or accessory also in his or her native natural
language (Spec. 1).
Independent claim 11, reproduced below, is representative of the
subject matter on appeal.
11. A method of performing recognition of a speech utterance from a
user with a distributed client- server system comprising the steps of:
(a) receiving user speech data from a client device in streaming
packets through a network interface of a network server system, said speech
data resulting from a first set of speech recognition operations being
performed on the speech utterance by a client device;
(b) recognizing the speech utterance as well as a native language used
in said speech utterance using processing routines executing at said network
server system which implement a second set of speech recognition
operations;
wherein said native language is identified based on identifying a
meaning of all of said speech utterance, and only after analyzing a combined
set of words and phrases presented therein;
(c) providing a response to the user in a same native language as
recognized in step (b);
wherein said processing routines at the network server system use one
or more speech recognition models that are trained and optimized based on
speech characteristics of a group of persons residing in geographical regions
served by the distributed client- server system.

REFERENCES
Harless US 5,730,603 Mar. 24, 1998
Barclay US 5,960,399 Sep. 28, 1999
Van Tichelen US 6,311,159 B1 Oct. 30, 2001
Appeal 2009-013176
Application 11/030,919

3
Joost US 6,327,568 B1 Dec. 4, 2001
Majaniemi US 6,393,403 B1 May 21, 2002
Kanevsky US 6,665,644 B1 Dec. 16, 2003

REJECTIONS
Claims 2-4, 9-11, 15, 22, 23, 28-30, 35, and 36 stand rejected under
35 U.S.C. § 103(a) as being unpatentable over Barclay in view of Kanevsky.
Claims 5, 7, 24, and 26 stand rejected under 35 U.S.C. § 103(a) as
being unpatentable over Barclay in view of Kanevsky and Van Tichelen.
Claims 8, 12, 27, and 31 stand rejected under 35 U.S.C. § 103(a) as
being unpatentable over Barclay in view of Kanevsky and Majaniemi.
Claims 6 and 25 stand rejected under 35 U.S.C. § 103(a) as being
unpatentable over Barclay in view of Kanevsky and Harless.
Claims 13, 14, 21, 32, and 33 stand rejected under 35 U.S.C. § 103(a)
as being unpatentable over Barclay in view of Kanevsky and Joost.

ISSUES
With respect to independent claims 11 and 30, Appellants argue that
Kanevsky does not show identifying native languages, but only dialects of
the same language (App. Br. 7). Appellants further contend that Kanevsky
does not show consideration of the entire utterance when determining native
language (App. Br. 9). Last, Appellants assert that there is no motivation to
combine Barclay with Kanevsky to obtain the invention under appeal (App.
Br. 10-11).
With respect to claims 5, 7, 24, and 26, Appellants argue that the
references do not teach or suggest an interactive electronic agent presented
within a browser or graphical agent of the client system (App. Br. 11).
Appeal 2009-013176
Application 11/030,919

4
With respect to claims 8, 12, 27, and 31, Appellants argue that
Majaniemi does not teach support for multiple languages in a portable cell
phone, as the Examiner suggests (App. Br. 11).
With respect to claims 13, 14, 21, 32, and 33, Appellants contend that
Joost should not be considered germane to the invention under appeal,
because the Examiner in a related application concluded that Joost is not
germane to distributed recognition environments (App. Br. 12).
Appellants’ arguments present us with the following issues:
1. Does Kanevsky teach identifying native language?
2. Does Kanevsky teach native language identification based on
identifying a meaning of all of the speech utterance?
3. Has the Examiner stated a motivation to combine Barclay and
Kanevsky having a rational underpinning?
4. Does the combination of Barclay, Kanevsky, and Van Tichelen
teach or fairly suggest an interactive electronic agent presented within a
browser or graphical agent of the client system?
5. Does the combination of Barclay, Kanevsky, and Majaniemi teach
or fairly suggest using a plurality of Hidden Markov Models to recognize the
speech utterance?
6. Should Joost be considered to be not germane to the claims of the
invention under appeal because an Examiner concluded that it was not
germane to the claims of a related application?

PRINCIPLES OF LAW
Section 103(a) forbids issuance of a patent when “the
differences between the subject matter sought to be patented
and the prior art are such that the subject matter as a whole
Appeal 2009-013176
Application 11/030,919

5
would have been obvious at the time the invention was made to
a person having ordinary skill in the art to which said subject
matter pertains.”
KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 406 (2007). The question of
obviousness is resolved on the basis of underlying factual determinations
including (1) the scope and content of the prior art, (2) any differences
between the claimed subject matter and the prior art, (3) the level of skill in
the art, and (4) where in evidence, so-called secondary considerations.
Graham v. John Deere Co., 383 U.S. 1, 17-18 (1966). See also KSR, 550
U.S. at 407 (“While the sequence of these questions might be reordered in
any particular case, the [Graham] factors continue to define the inquiry that
controls.”).
Absent a showing of good cause, the Board is not required to address
an argument in the Reply Brief that could have been presented in the
principal Brief. See Ex parte Borden, 93 USPQ2d 1473, 1473-74 (BPAI
2010) (“informative”3).

ANALYSIS
CLAIMS 2-4, 9-11, 15, 22, 23, 28-30, 35, AND 36
Appellants’ argument that Kanevsky does not show identifying native
languages, only dialects of the same language (App. Br. 7) is not persuasive.
The Examiner finds, and we agree, that Kanevsky teaches identifying native
language, as well as dialect, of a speaker. The acoustic front end of
Kanevsky is configured to extract at least one acoustic feature which is
correlated with at least one user attribute, including native language of the

3 The “informative” status of this opinion is noted at the following Board
website: http://www.uspto.gov/ip/boards/bpai/decisions/inform/index.jsp.
Appeal 2009-013176
Application 11/030,919

6
user (col. 3, ll. 51-57). “Native language of the user can be determined in a
manner essentially equivalent to accent classification” (col. 4, ll. 20-22).
Appellants’ argument that Kanevsky does not show consideration of
the entire utterance when identifying native language (App. Br. 9) is also not
persuasive. Appellants contend this is so because Kanevsky’s section on
determining dialect examines “keywords or idioms” rather than an entire
utterance (id.; Kanevsky col. 4, ll. 31-51). However, an argument that less
than the entire utterance is considered when determining dialect is not an
argument that Kanevsky fails to consider the entire utterance when
determining native language. Appellants thus fail to provide factual support
for this allegation.
We further disagree with Appellants’ contention that motivation to
combine Barclay with Kanevsky is absent from the references (App. Br. 10-
11). We find that the Examiner’s statements that the proposed combination
would have been obvious “for a purpose of modifying a behavior of a voice
system based on a user attribute for data mining” (Ans. 9) and that
“[t]ailoring the business logic to an attribute of a particular user through data
mining makes it more likely to sell a product to a user that is more
vulnerable to a proposal to buy a given product or service” (id. (citing
Kanevsky col. 7, ll. 48-54)) express a rationale for combining the references
having a rational underpinning.
Appellants’ argument that Kanevsky does not provide a response to
the user in a same native language as recognized (Reply Br. 6) is not entitled
to consideration, as the argument was not made in the principal Brief and no
showing of good cause for its late presentation has been made. See Borden,
supra.
Appeal 2009-013176
Application 11/030,919

7
We conclude that the Examiner did not in rejecting claims 2-4, 9-11,
15, 22, 23, 28-30, 35, and 36 as unpatentable over Barclay in view of
Kanevsky. We will sustain the Examiner’s § 103 rejection.
CLAIMS 5, 7, 24, AND 26
Appellants’ argument that there is no mention in the references
concerning how to incorporate an artificial conversational agent within a
web browser (App. Br. 11) is not persuasive. We agree with the Examiner’s
finding that Barclay teaches such a browser, and Van Tichelen teaches a
graphical user interface having an artificial conversational agent that is
controlled by speech (Ans. 19).
We conclude that the Examiner did not err in rejecting claims 5, 7, 24,
and 26 as unpatentable over Barclay in view of Kanevsky and Van Tichelen.
We will sustain the Examiner’s § 103 rejection.
CLAIMS 8, 12, 27, AND 31
Claims 8 and 27 recite, in pertinent part, that “a plurality of Hidden
Markov Models are used to recognize said speech utterance and said native
language . . . such that the system supports multiple native languages.”
Appellants’ argument that Majaniemi does not show support for
multiple languages in a portable cell phone and that Majaniemi does not
download multiple templates to support multiple languages (App. Br. 11) is
not persuasive to establish Examiner error. First, we agree with the
Examiner that Kanevsky is relied upon to teach automatic native language
identification (Ans. 21; Kanevsky col. 3, ll. 51-59; col. 4, ll. 25-31). Second,
contrary to Appellants’ argument, claims 8 and 27 do not require
simultaneous support of multiple languages, or automatic selection of a
Appeal 2009-013176
Application 11/030,919

8
particular language. Third, we agree with the Examiner’s finding that
Majaniemi teaches a plurality of Hidden Markov Models (col. 5, ll. 7-34).
Because we conclude that the Examiner’s combination is not
deficient, we will sustain the § 103 rejection of claims 8, 12, 27, and 31 over
Barclay in view of Kanevsky and Majaniemi.
CLAIMS 6 AND 25
Appellants rely on the arguments made with respect to parent claims
11 and 30. We will sustain the § 103 rejection of claims 6 and 25, then, for
the same reasons expressed supra with respect to claims 11 and 30.
CLAIMS 13, 14, 21, 32, AND 33
Appellants argue that Joost cannot be applied against these claims
because the Examiner, in another, “related” application,4 stated that Joost
was not germane to distributed recognition environments (App. Br. 12).
We are not persuaded by Appellants’ argument. We agree with the
Examiner that we need not conclude that Joost is not germane to this
application, because the claims in the two applications are not commensurate
in scope (Ans. 21).
We conclude that the Examiner did not err in rejecting claims 13, 14,
21, 32, and 33 under § 103 as being unpatentable over Barclay in view of
Kanevsky and Joost. We will sustain the Examiner’s rejection.

CONCLUSIONS
1. Kanevsky teaches identifying native language.
2. Kanevsky teaches native language identification based on
identifying a meaning of all of the speech utterance.

4 Application No. 11/030,864.
Appeal 2009-013176
Application 11/030,919

9
3. The Examiner has stated a motivation to combine Barclay and
Kanevsky having a rational underpinning.
4. The combination of Barclay, Kanevsky, and Van Tichelen teaches
an interactive electronic agent presented within a browser or graphical agent
of the client system.
5. The combination of Barclay, Kanevsky, and Majaniemi teaches
using a plurality of Hidden Markov Models to recognize the speech
utterance.
6. Joost should not be considered to be non-germane to the claims of
the invention under appeal simply because an Examiner concluded that it
was not germane to the claims of a related application, because the claims of
the two applications differ in scope.

DECISION
The Examiner’s decision rejecting claims 2-15, 21-33, 35, and 36 is
affirmed.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv) (2010).

AFFIRMED

babc