Ex Parte Kalthoff et al

Patent Trial and Appeal Board

Dec 2, 2013

11026427 (P.T.A.B. Dec. 2, 2013)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________________

Ex parte WOLFGANG KALTHOFF, JENS STAECK,
KARL FUERST, VOLKER SCHOTT, PETER LANG,
and MANFRED WALTER
____________________

Appeal 2011-004693
Application 11/026,427
Technology Center 2100
____________________

Before CAROLYN D. THOMAS, ELENI MANTIS MERCADER, and
JUSTIN BUSCH, Administrative Patent Judges.

BUSCH, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2011-004693
Application 11/026,427

2
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1-10, 12-25, and 27-31. We have jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.

Introduction
According to Appellants, the invention relates to determining data
object association based on graph theory techniques. Abstract; Spec. 1:3-23.

STATEMENT OF THE CASE
Exemplary Claim
Claim 1, reproduced below, is illustrative of the claimed subject
matter:
1. A method for determining associations between data
objects, the method comprising:
for a group of data objects including a first data object, a
second data object, and a third data object, executing a
matching algorithm to determine one or more duplicate data
object propositions among the data objects, a duplicate data
object proposition between two objects indicating a possibility
that the two objects are duplicates of each other;
detecting a duplicate data object proposition between the
first data object and the second data object;
identifying the first data object as a potential duplicate of
the second data object based on the duplicate data object
proposition between the first data object and the second data
object; and
in response to the identifying of the first data object as a
potential duplicate of the second data object, determining
whether the first data object identified as a potential duplicate
of the second data object is an actual duplicate of the second
data object by examining one or more duplicate data object
propositions including at least a duplicate data object
Appeal 2011-004693
Application 11/026,427

3
proposition between the third data object and the first data
object.

References
Budinsky
Pugh
US 6,407,753 B1
US 6,658,423 B1
Jun. 18, 2002
Dec. 2, 2003
Cooper
Kalthoff
US 2003/0172066 A1
US 2004/0103103 A1
Sep. 11, 2003
May 27, 2004

Rejections
Claims 1, 3-10, 12-16, 19-25, 27, 30, and 31 stand rejected under
35 U.S.C. §103(a) as being unpatentable over Cooper and Pugh. Ans. 4-19.
Claim 2 stands rejected under 35 U.S.C. §103(a) as being
unpatentable over Cooper, Pugh, and Kalthoff. Ans. 19-21.
Claims 17, 18, 28, and 29 stand rejected under 35 U.S.C. §103(a) as
being unpatentable over Cooper, Pugh, and Budinsky. Ans. 21-22.

ANALYSIS
Appellants argue the combination of Cooper and Pugh does not teach
“determining whether the first data object identified as a potential duplicate
of the second data object is an actual duplicate of the second data object by
examining one or more duplicate data object propositions including at least a
duplicate data object proposition between the third data object and the first
data object,” as recited in independent claim 1 and 19. App. Br. 9-12; Reply
1-3.
The Examiner finds Cooper teaches “determining whether the first
data object identified as a potential duplicate of the second data object is an
actual duplicate of the second data object.” Ans. 5-6, 26-27. In particular,
Appeal 2011-004693
Application 11/026,427

4
the Examiner finds Cooper initially determines whether two document
objects might be matches based on a comparison of the size of the objects
and subsequently determines whether the two objects are identical by
comparing terms in the document. Id. The Examiner finds that Pugh
teaches identifying near-duplicate documents by comparing each of the data
objects to a third data object. Ans. 6-7, 27-28 (describing Pugh’s method of
document clustering). The Examiner then finds it would be obvious to
combine Pugh’s method of determining near duplicates with Cooper because
Pugh allows organization of duplicate objects in Cooper into clusters “and
would reduce the direct comparisons necessary to determine duplicity in
Cooper’s system where a large number of duplicates exist.” Ans. 27.
Appellants argue the Examiner has broken apart a limitation, such that
the dependency within the limitation is read out of the claim. App. Br. 11-
12. Specifically, Appellants argue “the Examiner analyzes an operation of
determining whether the first data object is a potential duplicate of the
second data object [by examining another pair of objects].” App. Br. 11
(alteration in original). Appellants assert this analysis is improper because
“[d]etermining whether two objects are potential duplicates” is not the same
as determining actual duplicates of two potentially duplicate objects by
examining another pair of objects. Id. Appellants explain that the Examiner
merely finds Cooper teaches finding actual duplicates by a method not
recited (i.e., by comparing terms in a document) and Pugh teaches
identifying potential duplicates through its document clustering method.
App. Br. 11-12; Reply 2. Appellants further argue Pugh does not provide
any “hint or indication that the clustering approach . . . can be relied on for
determining actual duplicates.” Reply 2. Appellants thus conclude that the
Appeal 2011-004693
Application 11/026,427

5
Examiner’s combination of Cooper and Pugh does not teach determining
whether two objects are actual duplicates of each other by comparing one of
the two objects with a third object. App. Br. 12; Reply 2.
We agree with Appellants’ contention. While we agree with the
Examiner’s findings that Cooper teaches detecting and identifying potential
duplicates and subsequently determining whether two objects are actual
duplicates (by comparing document terms) and that Pugh teaches
determining potential duplicates by clustering, we do not find the
combination teaches or suggests determining whether the two objects are
actual duplicates by comparing at least one of the two objects with a third
object. The Examiner’s finding that it would be obvious to combine Pugh
with Cooper to reduce the number of direct comparisons does not provide a
combination that uses Pugh’s method to determine actual duplicates. In
particular, the Examiner’s motivation would result in using Pugh’s method
to identify clusters of near duplicates and subsequently use Cooper’s term
comparison method to compare the objects within a cluster to identify actual
duplicate objects. We do not agree that Pugh could be used to modify
Cooper in order to identify actual duplicates by Pugh’s method, particularly
in view of Pugh’s statement that:
it will be presumed that detecting near-duplicate document will
necessarily also detect exact duplicate documents. Therefore,
when the term “near-duplicate detection” is used, it will be
understood that exact duplicates will also be detected, though
not necessarily identified as “exact”, as opposed to near,
duplicates.
Pugh col. 7 ll. 39-44.
Appeal 2011-004693
Application 11/026,427

6
For the reasons as discussed above, we cannot sustain the rejection of
independent claims 1 and 19 or the rejection of dependent claims 2-10, 12-
18, 20-25, and 27-31, for similar reasons.

DECISION
For the above reasons, the Examiner’s rejection of claims 1-10, 12-25,
and 27-31 is reversed.

REVERSED

msc