Ex Parte Chou et al

Patent Trial and Appeal Board

Jul 15, 2014

10814081 (P.T.A.B. Jul. 15, 2014)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte WU CHOU, LI LI, and FENG LIU
____________

Appeal 2012-000757
Application 10/814,081
Technology Center 2600
____________

Before ST. JOHN COURTENAY III, THU A. DANG, and
LARRY J. HUME, Administrative Patent Judges.

COURTENAY, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1, 3, 5–7, and 9–18. Claims 2, 4, and 8 were cancelled. We have
jurisdiction under 35 U.S.C. § 6(b).
We affirm.

STATEMENT OF THE CASE
Appellants’ claimed invention is directed to language processing
techniques for the purpose of routing incoming calls in a communications
Appeal 2012-000757
Application 10/814,081

2

system (Spec. 1, ll. 5–17). Independent claim 1, reproduced below, is
representative of the subject matter on appeal:
1. A method comprising:

identifying, by a processor-based device, a word that (i)
is received in a communication, and (ii) is a natural-language
word, wherein the processor-based device is to determine a
category for the communication;

combining (i) at least one set of word-terms, and (ii) at
least one set of word-classes that is generated from an
automatic word-class clustering algorithm, wherein a term is
one of a word-term and a word-class, and wherein the
combining results in a combination of terms;

selecting by the processor-based device a plurality of
terms from the combination of terms, wherein the selecting is
based on an information-gain value of those terms in the
combination that correspond to the word;

generating by the processor-based device a matrix that
comprises a plurality of categories and the plurality of terms,
wherein each term in the matrix is associated with at least one
category; and

determining from the matrix, based on a joint
classification of the word by the processor-based device, the
category for the communication.

(Contested limitations emphasized).

REJECTION
Claims 1, 3, 5–7, and 9–18 stand rejected under 35 U.S.C. § 103(a) in
view of the combined teachings and suggestions of “Improving Latent
Semantic Indexing Based Classifier with Information Gain” (Li et al.,
Appeal 2012-000757
Application 10/814,081

3

published Sept. 2002), “An Unsupervised Method for Word Sense Tagging
Using Parallel Corpora” (Diab et al., published July 2002), and U.S. Patent
7,099,819 B2 (Sakai, filing date July 18, 2001).

GROUPING OF CLAIMS
Based on Appellants’ arguments, we decide the appeal of the § 103
rejection on the basis of representative claim 1 (App. Br. 16–20). See
37 C.F.R. § 41.37(c)(1)(vii) (2004).1

Independent Claim 1
Issue: Under § 103, did the Examiner err in finding the cited
combination of references would have taught or suggested the contested
limitations: “combining (i) at least one set of word-terms, and (ii) at least
one set of word-classes that is generated from an automatic word-class
clustering algorithm,” and “selecting . . . a plurality of terms from the
combination of terms, wherein the selecting is based on an information-gain
value of those terms in the combination that correspond to the word,” within
the meaning of representative claim 1?

Appellants’ principally contend the cited Diab reference “does not
teach the claimed combination of terms, because Diab associates to one
source of words (French words) possible senses that are deduced and

1 Appellants filed a Notice of Appeal on March 7, 2011. The date of filing
the Notice of Appeal determines which set of rules applies to an ex parte
appeal. If a Notice of Appeal is filed prior to January 23, 2012, then the
2004 version of the Board Rules last published in the 2011 edition of Title
37 of the Code of Federal Regulations (37 C.F.R. § 41.1) applies to the
appeal. See also MPEP 1220, Rev. 8, July 2010.
Appeal 2012-000757
Application 10/814,081

4

collected from an English corpus that comprises translations of French
words (Diab, page 256, col. 2 through page 257 col. 1)” (App. Br. 18).
Appellants urge “[t]he Specification teaches a distinctive combination of at
least two different kinds of sources of terms” (App. Br. 16).
In support of the rejection, the Examiner notes that two different kinds
of sources of terms are not claimed: “The relevant claim language is
‘combining (i) at least one set of word-terms and (ii) at least one set of word-
classes . . . wherein the combining results in a combination of terms’” (Ans.
36 (emphasis added)).
We agree with the Examiner (Ans. 35–36) the claims do not
specifically recite the argued feature of “different kinds of sources of terms”
(App. Br. 16). Our reviewing court guides: “In the patentability context,
claims are to be given their broadest reasonable interpretations . . . .
[L]imitations are not to be read into the claims from the specification.” In re
Van Geuns, 988 F.2d 1181, 1184 (Fed. Cir. 1993) (citations omitted).
Moreover, because “applicants may amend claims to narrow their scope, a
broad construction during prosecution creates no unfairness to the applicant
or patentee.” In re ICON Health and Fitness, Inc., 496 F.3d 1374, 1379
(Fed. Cir. 2007) (citation omitted).
Nevertheless, we consider the embodiments and associated
descriptions in the Specification for context. Appellants describe the
claimed “word-classes” as follows:
NLP [Natural Language Processing] generally involves
forming word term classes by clustering word terms that have
some common properties or similar semantic meanings. Such
word term classes are also referred to herein as “word classes,”
“clusters” or “classes.”
Appeal 2012-000757
Application 10/814,081

5

(Spec. 2, ll. 3–5 (emphasis added)).
The Examiner broadly interprets the claim term “word-classes” as
including word terms that have some common properties or similar semantic
meanings, and finds the claimed “word-classes” are taught are suggested by
Diab’s “senses” of words that are sets of words which have similar
meanings. (Ans. 30–31). Given Appellants’ exemplary non-limiting
description supra for “word classes” (Spec. 2), on this record, we are not
persuaded the Examiner’s interpretation is overly broad, unreasonable, or
inconsistent with the Appellants’ Specification.
Appellants further contend Diab does not teach “the IG-based analysis
and selection recited in claim 1.” (App. Br. 19).
In support of the rejection, the Examiner notes:
the rejection is based on a combination . . . . Since Li teaches
the selection of both types of information, it does not matter
that Diab does not teach selection of both types of information
because Li addresses the selection limitation and Diab was only
applied to teach the characteristic of the word/class information
source.
Appellant’s arguments directed only to a single reference
do not overcome a rejection when the rejection is based on a
combination.
(Ans. 41 (emphasis omitted)).
Our reviewing court guides that one cannot show nonobviousness by
attacking references individually where the rejection is based on a
combination of references. In re Merck & Co., Inc., 800 F.2d 1091, 1097
(Fed. Cir. 1986). Therefore, we do not find persuasive Appellants’ singular
attack on Diab. (App. Br. 19).
Appeal 2012-000757
Application 10/814,081

6

Appellants further contend the Examiner’s combination of the Li and
Diab references is improper. Appellants contend, inter alia, “nowhere does
Li-2002 in any way suggest a combination of, or a replacement with
different kinds of, corpora.” (App. Br. 19).
The Examiner’s proffered motivation for combination is that “one of
ordinary skill in the art can simply substitute the corpus used in Li with one
generated from a processor as per Diab that contains the information needed
to generate the matrix, and which can be used to bootstrap the classifier
described in Li.” (Ans. 8).
We find the Examiner has provided sufficient articulated reasoning
with some rational underpinning to support the legal conclusion of
obviousness. (Ans. 8). While we are fully aware that hindsight bias often
plagues determinations of obviousness, Graham v. John Deere Co., 383
U.S. 1, 36 (1966), we are also mindful the Supreme Court has clearly stated
the “combination of familiar elements according to known methods is likely
to be obvious when it does no more than yield predictable results.” KSR
Int’l Co. v. Teleflex Inc., 550 U.S. 398, 416 (2007). We are further mindful
the skilled artisan would “be able to fit the teachings of multiple patents
together like pieces of a puzzle” (KSR, 550 U.S. at 420) since the skilled
artisan is “a person of ordinary creativity, not an automaton.” (Id. at 21).
Here, we are not persuaded the Examiner’s proffered combination of
would have been “uniquely challenging or difficult for one of ordinary skill
in the art.” See Leapfrog Enters., Inc. v. Fisher-Price, Inc., 485 F.3d 1157,
1162 (Fed. Cir. 2007) (citing KSR, 550 U.S. at 418). Therefore, on this
record, we are not persuaded the Examiner erred by improperly combining
the references under §103. (Ans. 8).
Appeal 2012-000757
Application 10/814,081

7

Accordingly, we sustain the Examiner’s § 103 rejection of
representative claim 1, and grouped claims 3, 5–7, and 9–18 (not argued
separately), which fall therewith. See 37 C.F.R. § 41.37(c)(1)(vii) (2004).
DECISION
We affirm the Examiner’s § 103 rejection of claims 1, 3, 5–7,
and 9–18.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 41.50(f).
AFFIRMED

msc