Ex Parte Lakshminarayan et al

Patent Trial and Appeal Board

Oct 22, 2012

10871148 (P.T.A.B. Oct. 22, 2012)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte CHOUDUR K. LAKSHMINARAYAN,
PRAMOD SINGH, and QINGFENG YU
____________

Appeal 2010-006593
Application 10/871,148
Technology Center 2100
____________

Before JOHN A. JEFFERY, JENNIFER L. McKEOWN,
and DAVID C. McKONE, Administrative Patent Judges.

McKONE, Administrative Patent Judge.

DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1-9 and 18-23, which constitute all the claims pending in this
application. See Apr. 23, 2009, Notice of Appeal. Claims 10-17 were
withdrawn in response to a restriction requirement. We have jurisdiction
under 35 U.S.C. § 6(b).
We affirm.

Appeal 2010-006593
Application 10/871,148

2
STATEMENT OF THE CASE
Appellants’ invention relates to clustering and classification of data in
data mining. See, e.g., Spec. 3-4. Claim 1, which is illustrative of the
invention, reads as follows:
1. A method for grouping a plurality of
data elements of a dataset, comprising:
clustering said dataset into a plurality of
clusters, each of said plurality of clusters
comprising at least one of said plurality of data
elements; and
iteratively classifying said plurality of
clusters, separately from the clustering of said
dataset, into a plurality of classes of like data
elements.

THE REJECTIONS
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Taylor US 2002/0165839 A1 Nov. 7, 2002
Hubey US 2003/0065632 A1 Apr. 3, 2003
Chen US 2004/0220963 A1 Nov. 4, 2004
(filed May 1, 2003)
Chilingarian US 2006/0172292 A1 Aug. 3, 2006
(filed Feb. 28, 2003)
Hsiung US 7,136,716 B2 Nov. 14, 2006
(division of 09/802,519, filed Mar. 9, 2001)

1. Claims 1, 2, 7, 19-21, and 23 stand rejected under 35 U.S.C.
§ 103(a) as unpatentable over Hubey and Chen. See Ans. 3-7.1

1 Throughout this opinion, we refer to the Appeal Brief filed August 31,
2009, and the Examiner’s Answer mailed November 17, 2009.
Appeal 2010-006593
Application 10/871,148

3
2. Claims 3-6 and 22 stand rejected under 35 U.S.C. § 103(a) as
unpatentable over Hubey, Chen, and Taylor. See Ans. 8-10.
3. Claims 8 and 9 stand rejected under 35 U.S.C. § 103(a) as
unpatentable over Hubey, Chen, and Chilingarian. See Ans. 10-12.
4. Claim 18 stands rejected under 35 U.S.C. § 103(a) as obvious
over Hubey, Chen, and Hsiung. See Ans. 12-13.

ISSUES
Although Appellants nominally argue independent claims 18, 19, and
23 separately from claim 1, see Br. 12-14, the arguments presented for
claims 18, 19, and 23 are substantially the same as those presented for claim
1. Appellants do not separately argue the rejections of the dependent claims.
See Br. 5, 9. Accordingly, we will discuss Appellants’ contentions by
reference to claim 1.
Appellants contend that:
1. Hubey does not disclose an “iteratively classifying [a] plurality
of clusters”;
2. Chen does not disclose “classifying said plurality of clusters . . .
into a plurality of classes of like data elements”; and
3. Hubey does not disclose “iteratively classifying [a] plurality of
clusters separate from the clustering of said dataset,”
as recited in claim 1 (emphasis added). See Br. 10-11.
Accordingly, the issues raised by Appellants’ contentions are as
follows: With regard to each of the above referenced claim limitations, did
the Examiner err in finding that the combination of Hubey and Chen teaches
or suggests the limitation?
Appeal 2010-006593
Application 10/871,148

4
ANALYSIS
Rejection of Claims 1, 2, 7, 19-21, and 23 under 35 U.S.C. § 103(a)
Appellants contend that Hubey does not disclose “iteratively
classifying [a] plurality of clusters.” Br. 11. The Examiner found that
Hubey, at Paragraph 0007, discloses this limitation. See Ans. 4. We agree
with Appellants. While Hubey discloses classification, see Hubey at ¶ 0010,
that classification is not described as iterative.
That finding is not determinative, however, because the Examiner also
found that Chen’s disclosure of an “iterative projecting technique for
clustering” constitutes “iteratively classifying.” See Ans. at 4-5 (citing Chen
¶¶ 0029-30). The Examiner explains that this iterative projecting technique
“enables separate clustering of objects that have separate data types to
contribute to the clustering process.” Ans. 16 (quoting Chen ¶ 0029). The
Examiner further notes that Chen’s “iterative clustering projection technique
relies on obtaining clustering information from separate types of objects that
are arranged in separate layers, with each layer containing a homogenous
type of object.” Ans. 16 (quoting Chen ¶ 0030).
Appellants contest this finding, arguing that “Chen is directed to an
iterative clustering technique,” which they distinguish from “iteratively
classifying a plurality of clusters into a plurality of classes of like data
elements.” Br. 11. After characterizing Chen as disclosing “clustering
different types of objects and using the information derived from the
clustering of one type of object in the clustering of the other type of object,”
Appellants simply argue that this is different from iteratively classifying. Id.
Appellants, however, do not explain why Chen’s “iterative clustering
Appeal 2010-006593
Application 10/871,148

5
projection technique” is distinguishable from the “iteratively classifying”
claimed in claim 1.
Although it gives examples of both, Appellants’ Specification does
not define either “clustering” or “classifying.” In the described examples,
“clustering methods are implemented in conjunction with classification
schemes . . . to provide more accurate classification of data elements than
could otherwise be obtained by traditional clustering algorithms alone.”
Spec. ¶ 0034. The Specification states that classification schemes “address
asymmetrical covariance structures in the clusters,” Spec. ¶ 0034, and gives
examples of classifying schemes, such as “statistical classification
procedure[s],” Spec. ¶ 0042, and “discriminant analysis techniques,” Spec. ¶
0046. However, nothing in the Specification suggests that “classifying” is
limited to the described examples.
Moreover, the Specification does not draw a clear line between
“classifying” and “clustering.” On one hand, both “clustering” and
“classifying” are equated with “groups” of data. See Spec. ¶¶ 0037 (“a
plurality of data elements or observations is grouped or clustered”);
(emphasis added), 0039 (“the grouping or class of data elements is
sufficiently homogenous”); (emphasis added), 0041 (“data elements are
clustered or grouped”). (emphasis added). This suggests that clusters and
classes may be the same thing, or at least overlap. This is consistent with
Hubey’s explanation, at ¶ 0010, that classification “is simply a form [of]
cluster[ing].”
On the other hand, the Specification describes “a two-stage clustering
and classification approach,” ¶ 0035, suggesting that “clustering” and
“classification” are distinct. However, determining when the clustering ends
Appeal 2010-006593
Application 10/871,148

6
and the classification begins is not so simple. For example, Appellants’
Specification at Paragraph [0036] describes a “clustering algorithm” or
“partitioning method” that decomposes a “super-cluster” into “component
groups.” In this method, the data undergo a “hierarchical divisive grouping
process” to divide the large cluster into smaller “groups.”
Those “groups” are later equated to “class[es]” (“To determine if a
group or class should be further divided . . . .”). Spec. ¶ 0036 (emphasis
added). Thus, Paragraph 0036 describes a clustering algorithm that divides a
super cluster into plural classes. In short, the Specification does not support
Appellants’ narrow reading of “classifying” in claim 1 to exclude Chen’s
“iterative projecting technique for clustering.”
“During examination, claims are to be given their broadest reasonable
interpretation consistent with the specification, and claim language should
be read in light of the specification as it would be interpreted by one of
ordinary skill in the art.” In re Am. Acad. of Sci. Tech Ctr., 367 F.3d 1359,
1364 (Fed. Cir. 2004) (internal quotation marks omitted). Under such an
interpretation, Chen’s iterative clustering technique can be considered
“classifying.” In Chen, different types of objects are clustered in different
“layers.” See Chen ¶ 0026. Chen gives the example of “clustering of
multiple types of objects in which each type of object is substantially
identical (e.g., one type pertains to a group of web pages, a group of users,
or a group of documents, etc.).” Chen ¶ 0027. In this example, web pages
would be clustered in a different layer than documents. See id. The result is
that “each layer contain[s] a homogenous type of object.” Chen ¶ 0030.
This is comparable to Appellants’ use of “a two-stage clustering and
classification approach to derive a homogeneous set of observations within
Appeal 2010-006593
Application 10/871,148

7
each cluster,” Spec. ¶ 0035. The Examiner pointed out that Chen likens its
clusters to classes, explaining that “object types (that are to be clustered
according to the clustering algorithm) can be considered as the instance of a
‘latent’ class.” Ans. 17 (quoting Chen ¶ 0029). Appellants do not explain
why the Examiner’s interpretation of classifying is unreasonably broad.
Accordingly, we agree with the Examiner’s finding that Chen describes
“classifying said plurality of clusters . . . into a plurality of classes of like
data elements.”
The Examiner also found that Chen describes this classification as
iterative. See Ans. 4-5 (citing Chen ¶¶ 0029-0030). While Appellants
contest whether Chen’s technique is classifying, they do not contest that it is
iterative. See Br. 11. Accordingly, we agree with the Examiner’s finding
that Chen discloses “iteratively classifying said plurality of clusters.”
(emphasis added).
Regarding the claim language “separately from the clustering of said
dataset,” the Examiner acknowledges that Hubey does not disclose iterative
classification separate from clustering. See Ans. 4. However, the Examiner
found that Chen’s “iterative projecting technique” can be combined with and
performed separately from Hubey’s clustering. Ans. 4-5. While Appellants
contend that Chen does not disclose “classifying,” they do not contest the
Examiner’s finding that Chen’s “iterative projecting technique” can be
performed in combination with, and separate from, Hubey’s clustering. As
explained above, we agree with the Examiner’s finding that Chen discloses
classifying; thus we agree with the Examiner’s finding that Hubey and Chen
together show classification “separately from the clustering of said dataset.”
Appeal 2010-006593
Application 10/871,148

8
Accordingly, we sustain the rejection of (1) independent claim 1; (2)
independent claims 19 and 23 which recite commensurate limitations; and
(3) dependent claims 2, 7, 20, and 21 not separately argued.2

Rejection of Claims 3-6, 8, 9, and 22 under 35 U.S.C. § 103(a)
Since Appellants present no arguments pertaining to the Examiner’s
obviousness rejections of claims 3-6, 8, 9, and 22, we summarily sustain
these rejections. See MPEP § 1205.02 (“If a ground of rejection stated by
the examiner is not addressed in the appellant’s brief, that ground of
rejection will be summarily sustained by the Board.”).

Rejection of Claim 18 under 35 U.S.C. § 103(a)
The Appellants’ arguments regarding claim 18 are substantially the
same as those presented for claim 1. See Br. 13-14. Accordingly, we

2 If prosecution is reopened, the Examiner should consider a rejection of
claim 1 under 35 U.S.C. § 101. It appears that claim 1 recites nothing more
than an algorithm that could be performed within one’s head. “Such a
method that can be performed by human thought alone is merely an abstract
idea and is not patent-eligible under § 101.” CyberSource Corp. v. Retail
Decisions, Inc., 654 F.3d 1366, 1373 (Fed. Cir. 2011). Rather the describing
a concrete application of an abstract idea, claim 1 would preempt in all fields
the idea of clustering and iteratively classifying the clusters separately from
the clustering. See Bilski v. Kappos, 130 S. Ct. 3218, 3231 (2010) (“The
concept of hedging, described in claim 1 and reduced to a mathematical
formula in claim 4, is an unpatentable abstract idea. . . . Allowing
petitioners to patent risk hedging would pre-empt use of this approach in all
fields, and would effectively grant a monopoly over an abstract idea.”).

Appeal 2010-006593
Application 10/871,148

9
sustain the Examiner’s rejection of claim 18 for the same reasons as given
for claim 1, above.
ORDER
The decision of the Examiner to reject claims 1-9 and 18-23 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). See 37 C.F.R.
§ 1.136(a)(1)(iv) (2010).

AFFIRMED

rwk