Ex Parte Bent

Patent Trial and Appeal Board

Oct 11, 2018

11550583 (P.T.A.B. Oct. 11, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
11/550,583 10/18/2006
46320 7590
Shutts & Bowen LLP
STEVEN M. GREENBERG
525 Okeechobee Blvd
# 1100
West Palm Beach, FL 33401
10/15/2018
FIRST NAMED INVENTOR
Graham Anthony Bent
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
GB920050069US1 (430) 8929
EXAMINER
RIFKIN, BEN M
ART UNIT PAPER NUMBER
2123
NOTIFICATION DATE DELIVERY MODE
10/15/2018 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):
docketing@crgolaw.com
sgreenberg@shutts.com
aschneider@shutts.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte GRAHAM ANTHONY BENT
Appeal2017-007597
Application 11/550,583 1
Technology Center 2100
Before HUBERT C. LORIN, ANTON W. PETTING, and
MICHAEL W. KIM, Administrative Patent Judges.
LORIN, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Graham Anthony Bent (Appellant) seeks our review under 35 U.S.C.
§ 134(a) of the final rejection of claims 1, 9, 15, 21, 24, 26, 28, 31, and 32.
We have jurisdiction under 35 U.S.C. § 6(b).
SUMMARY OF DECISION
We REVERSE.
1 The Appellant identifies International Business Machines Corporation as
the real party in interest. App. Br. 2.
Appeal2017-007597
Application 11/550,583
THE INVENTION
Claim 1, reproduced below, is illustrative of the subject matter on
appeal.
1. A computer-implemented method for constructing a classifier
classifying data records into specific categories, comprising:
a) clustering a set of records that are to be classified into a
plurality of clusters of records by applying a clustering process
to the set of records that are to be classified;
b) creating a first classifier based upon the clustered set of
records that are to be classified, the first classifier enabled to
classify records into a plurality of training clusters;
c) applying the first classifier to a set of training records to
create a plurality of training clusters, each record in each training
cluster of the plurality of training clusters having a corresponding
predicted class;
d) creating separate classifiers for each training cluster of
the plurality of training clusters; and,
e) applying one of the created separate classifiers to a
corresponding one of the clusters of records amongst the
clustered set of records in order to classify the corresponding one
of the clusters of records,
wherein the first classifier and also each separate classifier
learns one or more sets of relationships that define each of a
known set of predicted classes and wherein the computer-
implemented method first processes the set of records that are to
be classified to determine a context before using the set of
training records to classify the set of records based upon the
determined context.
2
Appeal2017-007597
Application 11/550,583
THE REJECTION
The Examiner relies upon the following as evidence of
unpatentability:
Vu
Goodman
US 2002/0078091 Al
US 2005/0015454 Al
June 20, 2002
Jan.20,2005
Ricardo Gutierrez-Osuna, Intelligent Sensor Systems - Lecture 13:
Validation, Wright State University (August 2, 2002). [Gutierrez-
Osuna]
The following rejection is before us for review:
Claims 1, 9, 15, 21, 24, 26, 28, 31, and 32 are rejected under
35 U.S.C. § I03(a) as being unpatentable over Goodman, Vu, and Gutierrez-
Osuna.
ISSUE
Did the Examiner err in rejecting claims 1, 9, 15, 21, 24, 26, 28, 31,
and 32 under 35 U.S.C. § I03(a) as being unpatentable over Goodman, Vu,
and Gutierrez-Osuna?
ANALYSIS
With respect to claim 1, the Examiner finds that Goodman discloses
all that is claimed but
fails to explicitly disclose, "b) creating a first classifier based
upon the set of records that are to be classified, the first classifier
enabled to classify records into a plurality of training clusters"
and "applying the first classifier", "e) applying one of the created
separate classifiers to a corresponding one of the clusters of
records amongst the clustered set of records in order to classify
3
Appeal2017-007597
Application 11/550,583
the corresponding one of the clusters of records", and "Wherein
the computer-implemented method first processes the set of
records that are to be classified to determine a context before
using the set of training records to classify the set of records
based upon the determined context."
Final Rej. 3--4. Vu is relied upon for disclosing step "b)" and Gutierrez-
Osuna is relied upon for disclosing step "e )". The Examiner makes similar
findings with respect to the other independent claims (See Final Rej. 6 and 9
for independent claims 9 and 15, respectively). According to the Examiner,
"[a]t the time of invention it would have been obvious to one skilled in the
art of clustering to combine the work of Goodman and Vu in order to"
• "make use of k nearest neighbor to cluster and to make use of context
when classifying" (Final Rej. 4, 8, and 11); and,
• "allow the user to test itself on its own training data" (Final Rej. 5, 9,
and 12).
The Appellant disagrees that Gutierrez-Osuna discloses step "e )".
The Examiner relied on page 3 of Gutierrez-Osuna as disclosing the
subject matter of claim step "(e)." Final Rej. 4, 8, and 11 ("(Pg.3; EN: this
denotes the process of using the training set to estimate the error rate (i.e.
classify the training data used to create the classifier).")
The Appellant reproduced page 3 of Gutierrez-Osuna and argues that
"[a]s can be seen, the portion referred to by Examiner teaches nothing
related to the application of a classifier to any set of records so as to classify
the records." App. Br. 6.
At best, page 3 of Gutierrez-Osuna discloses that, "in real
applications," one only has "access to a finite set of examples, usually
smaller than [ ] wanted." "One approach is to use the entire training data to
4
Appeal2017-007597
Application 11/550,583
select our classifier and estimate the error rate." "A much better approach is
to split the training data into disjoint subsets: the holdout method." But said
disclosure does not meet what is claimed, which is "[ e)] applying one of the
created separate classifiers to a corresponding one of the clusters of records
amongst the clustered set of records in order to classify the corresponding
one of the clusters of records." The claims do not simply call for splitting
data into subsets, but rather to apply a created separate classifier to a
corresponding cluster of records "in order to classify the corresponding one
of the clusters of records."
The Answer repeats the reliance on page 3 of Gutierrez-Osuna as
disclosing the subject matter of claim step "( e )" as repeated in the Answer.
Ans. 4. But, as the Appellant argues, "nothing is present in page 3 of
Gutierrez-Osuna of the application of a classifier to a corresponding cluster
of records in order to classify the corresponding cluster of records." Reply
Br. 5.
The rejection is not sustained.
DECISION
The decision of the Examiner to reject claims 1, 9, 15, 21, 24, 26, 28,
31, and 3 2 is reversed.
REVERSED
5