Ex Parte Busche

Board of Patent Appeals and Interferences

May 28, 2009

09879491 (B.P.A.I. May. 28, 2009)

UNITED STATES PATENT AND TRADEMARK OFFICE 1
___________ 2
3
BEFORE THE BOARD OF PATENT APPEALS 4
AND INTERFERENCES 5
___________ 6
7
Ex parte FREDERICK D. BUSCHE 8
___________ 9
10
Appeal 2008-004750 11
Application 09/879,491 12
Technology Center 3600 13
___________ 14
15
Decided:1 May 28, 2009 16
___________ 17
18
Before MURRIEL E. CRAWFORD, ANTON W. FETTING, and 19
BIBHU R. MOHANTY, Administrative Patent Judges. 20
21
FETTING, Administrative Patent Judge. 22
23
24
DECISION ON APPEAL 25
26
STATEMENT OF THE CASE27

1 The two month time period for filing an appeal or commencing a civil
action, as recited in 37 C.F.R. § 1.304, begins to run from the decided date
shown on this page of the decision. The time period does not run from the
Mail Date (paper delivery) or Notification Date (electronic delivery).

Appeal 2008-004750
Application 09/879,491

2
Frederick D. Busche (Appellant) seeks review under 35 U.S.C. § 134 1
of a non-final rejection of claims 1-8, 10-22, 24-35, and 37-43, the only 2
claims pending in the application on appeal. 3
We have jurisdiction over the appeal pursuant to 35 U.S.C. § 6(b) 4
(2002). 5
We AFFIRM-IN-PART. 6
The Appellant invented a way of predicting customer behavior based 7
on data network geography (Specification 3:6-8). 8
An understanding of the invention can be derived from a reading of 9
exemplary claims 1, 15, 29, and 41-43 which are reproduced below 10
[bracketed matter and some paragraphing added]. 11
1. 1. A data processing machine implemented method of 12
selecting data sets for use with a predictive algorithm based on 13
data network geographical information, comprising data 14
processing machine implemented steps of: 15
[1] generating a first statistical distribution of a training data 16
set; 17
[2] generating a second statistical distribution of a testing data 18
set; 19
[3] using the first statistical distribution and the second 20
statistical distribution to identify a discrepancy between the first 21
statistical distribution and the second statistical distribution 22
with respect to the data network geographical information by 23
comparing at least one of the first statistical distribution and the 24
second statistical distribution to a statistical distribution of a 25
customer database to determine if at least one of the training 26
data set and the testing data set are geographically 27
representative of a customer population represented by the 28
customer database; 29
[4] modifying selection of entries in one or more of the training 30
data set and the testing data set based on the discrepancy 31
Appeal 2008-004750
Application 09/879,491

3
between the first statistical distribution and the second 1
statistical distribution; and 2
[5] using the modified selection of entries by the predictive 3
algorithm. 4
5
15. An apparatus for selecting data sets for use with a predictive 6
algorithm based on data network geographical information, 7
comprising: 8
[1] a statistical engine; 9
[2] a comparison engine coupled to the statistical engine, 10
wherein the statistical engine 11
generates a first statistical distribution of a training data set and 12
a second distribution of a testing data set, 13
the comparison engine 14
uses the first statistical distribution and the second distribution 15
to identify a discrepancy between the first statistical distribution 16
and the second distribution with respect to the data network 17
geographical information by comparing at least one of the first 18
statistical distribution and the second statistical distribution to a 19
statistical distribution of a customer database to determine if at 20
least one of the training data set and the testing data set are 21
geographically representative of a customer population 22
represented by the customer database, 23
modifies selection of entries in one or more of the training data 24
set and the testing data set based on the discrepancy between 25
the first statistical distribution and the second distribution, and 26
provides the modified selection of entries for use by the 27
predictive algorithm; and 28
[3] a predictive algorithm device that uses the modified 29
selection of entries and the predictive algorithm. 30
31
29. A computer program product in a computer readable 32
medium comprising instructions for enabling a data processing 33
Appeal 2008-004750
Application 09/879,491

4
machine to select data sets for use with a predictive algorithm 1
based on data network geographical information, comprising: 2
[1] first instructions for generating a first statistical distribution 3
of a training data set; 4
[2] second instructions for generating a second statistical 5
distribution of a testing data set; 6
[3] third instructions for using the first statistical distribution 7
and the second statistical distribution to identify a discrepancy 8
between the first statistical distribution and the second 9
statistical distribution with respect to the data network 10
geographical information by comparing at least one of the first 11
statistical distribution and the second statistical distribution to a 12
statistical distribution of a customer database to determine if at 13
least one of the training data set and the testing data set are 14
geographically representative of a customer population 15
represented by the customer database; 16
[4] fourth instructions for modifying selection of entries in one 17
or more of the training data set and the testing data set based on 18
the discrepancy between the first statistical distribution and the 19
second statistical distribution; and 20
[5] fifth instructions for using the modified selection of entries 21
by the predictive algorithm. 22
23
41. A data processing machine implemented method of 24
predicting customer behavior based on data network 25
geographical influences, comprising data processing machine 26
implemented steps of: 27
[1] obtaining data network geographical information regarding 28
a plurality of customers, 29
the data network geographic information comprising frequency 30
distributions of both 31
(i) number of data network links between a customer 32
geographical location and one or more web site data network 33
geographical locations, and 34
Appeal 2008-004750
Application 09/879,491

5
(ii) size of a click stream for arriving at the one or more web 1
site data network geographical locations; 2
[2] training a predictive algorithm using the data network 3
geographical information; and 4
[3] using the predictive algorithm to predict customer behavior 5
based on the data network geographical information. 6
7
42. An apparatus for predicting customer behavior based on 8
data network geographical influences, comprising: 9
[1] means for obtaining data network geographical information 10
regarding a plurality of customers, the data network geographic 11
information comprising frequency distributions of both 12
(i) number of data network links between a customer 13
geographical location and one or more web site data network 14
geographical locations, and 15
(ii) size of a click stream for arriving at the one or more web 16
site data network geographical locations; 17
[2] means for training a predictive algorithm using the data 18
network geographical information; and 19
[3] means for using the predictive algorithm to predict customer 20
behavior based on the data network geographical information. 21
22
43. A computer program product in a computer readable 23
medium comprising instructions for enabling a data processing 24
machine to predict customer behavior based on data network 25
geographical influences, comprising: 26
[1] first instructions for obtaining data network geographical 27
information regarding a plurality of customers, the data network 28
geographic information comprising frequency distributions of 29
both 30
(i) number of data network links between a customer 31
geographical location and one or more web site data network 32
geographical locations, and 33
Appeal 2008-004750
Application 09/879,491

6
(ii) size of a click stream for arriving at the one or more web 1
site data network geographical locations; 2
[2] second instructions for training a predictive algorithm using 3
the data network geographical information; and 4
[3] third instructions for using the predictive algorithm to 5
predict customer behavior based on the data network 6
geographical information. 7
8
This appeal arises from the Examiner’s Non-Final Rejection, mailed 9
June 1, 2007. The Appellant filed an Appeal Brief in support of the appeal 10
on October 31, 2007. An Examiner’s Answer to the Appeal Brief was 11
mailed on January 16, 2008. 12
13
PRIOR ART 14
The Examiner relies upon the following prior art: 15
Menon US 5,537,488 Jul. 16, 1996 16
Wu US 6,741,967 B1 May 25, 2004 17
18
REJECTIONS 19
Claims 1-8, 10-22, 24-35, and 37-43 stand rejected under 35 U.S.C. § 20
101 as directed to non-statutory subject matter. 21
Claims 1, 15, 29, and 41-43 stand rejected under 35 U.S.C. § 112, first 22
paragraph, as lacking a supporting written description within the original 23
disclosure. 24
Appeal 2008-004750
Application 09/879,491

7
Claims 1-8, 10-22, 24-35, and 37-43 stand rejected under 35 U.S.C. § 1
103(a) as unpatentable over Menon, Wu, and Appellant’s Admitted Prior 2
Art.2 3
4
ISSUES 5
The issue of whether the Appellant has sustained its burden of 6
showing that the Examiner erred in rejecting claims 1-8, 10-22, 24-35, and 7
37-43 rejected under 35 U.S.C. § 101 as directed to non-statutory subject 8
matter turns on the category of subject matter and the machine or 9
transformation test. 10
The issue of whether the Appellant has sustained its burden of 11
showing that the Examiner erred in rejecting claims 1, 15, 29, and 41-43 12
under 35 U.S.C. § 112, first paragraph, as lacking a supporting written 13
description within the original disclosure turns primarily on what is meant 14
by “using.” 15
The issue of whether the Appellant has sustained its burden of 16
showing that the Examiner erred in rejecting claims 1-8, 10-22, 24-35, and 17
37-43 under 35 U.S.C. § 103(a) as unpatentable over Menon, Wu, and 18
Appellant’s Admitted Prior Art turns primarily on whether Wu describes the 19
particular type of analysis claimed. 20
21

2 The Examiner couched this rejection as two separate rejections with the
order of the references cited changed in each rejection. We combine these
rejections for administrative convenience given that the claims are rejected
over the same art in each case.
Appeal 2008-004750
Application 09/879,491

8
FACTS PERTINENT TO THE ISSUES 1
The following enumerated Findings of Fact (FF) are supported by a 2
preponderance of the evidence. 3
Facts Related to Claim Construction 4
01. The Specification defines data network geography as the 5
collective morass of web sites and web pages that make up the 6
data network navigated to ultimately arrive at the goods and 7
services that customers wish to purchase (Specification 4:16-20). 8
Facts Related to Appellant’s Disclosure 9
02. The Specification describes the state of the art at the time of 10
filing as such that, when using artificial intelligence algorithms to 11
discover patterns in behavior exhibited by customers, it is 12
necessary to create training data sets where a predicted outcome is 13
known as well as testing data sets where the predicted outcome is 14
known to be able to validate the accuracy of a predictive 15
algorithm. The predictive algorithm, for example, may be 16
designed to predict a customer's propensity to respond to an offer 17
or his propensity to buy a product. It was also known that ease of 18
access to various goods and services may also influence the 19
customer's ultimate purchase patterns. That is, if a customer is 20
able to obtain access to the goods and services more easily, the 21
customer is typically more likely to engage in the purchase of such 22
goods and services (Specification 3:10 – 4:13). 23
03. The Specification describes using the modified selection of 24
entries by a predictive algorithm. The Specification states that if a 25
user inputs request parameters for requesting a prediction of 26
Appeal 2008-004750
Application 09/879,491

9
customer behavior, the customer behavior rules will be applied to 1
the input parameters and a customer behavior prediction will be 2
output (Specification 45:16-20). 3
Menon 4
04. Menon is directed to pattern recognition for recognizing input 5
data patterns from a subject and classifying the subject. Menon 6
first performs a training operation in which input training patterns 7
are received and grouped into clusters. Each cluster of training 8
patterns is associated with a category having a category definition 9
based on the training patterns in the cluster. As each training 10
pattern is received, a correlation or distance is computed between 11
it and each of the existing categories. Based on the correlations, a 12
best match category is selected. The best match correlation is 13
compared to a preset training correlation threshold. If the 14
correlation is above the threshold, then the training pattern is 15
added to the cluster of the best match category, and the definition 16
of the category is updated in accordance with a learning rule to 17
include the contribution from the new training pattern. If the 18
correlation is below the threshold, a new category defined by the 19
training pattern is formed, the cluster of the new category having 20
only the single training pattern (Menon 1:22-40). 21
05. To label categories, Menon counts the number of training 22
patterns of each class within the pattern cluster of each category. 23
It uses the counts to generate a training class histogram for each 24
category which shows the number of training patterns of each 25
class within the category's cluster. Menon uses the training 26
Appeal 2008-004750
Application 09/879,491

10
histograms of the categories to assign labels to the categories 1
(Menon 2:21-42). 2
06. Menon combines the learning features of adaptive pattern 3
recognition systems such as neural networks with statistical 4
decision making to perform its classifications. The definition of 5
categories during training, the labeling of the categories and the 6
output classifications are all performed in terms of histograms. 7
Thus, the classifications are associated with a probability of 8
correct classification (Menon 4:14-21). 9
07. When Menon’s system is trained, it receives training data 10
patterns from various subjects or classes. Each training pattern is 11
associated with a known class and takes the form of a feature 12
pattern vector IINP. Each category definition Ik is expressed in a 13
vector format compatible with the feature vector. As each pattern 14
vector is received, a correlation CTRN between it and each existing 15
category definition is performed. The correlation CTRN is then 16
compared to a preset training threshold λTRN. If a category is 17
found for which the correlation CTRN exceeds the threshold λTRN, 18
then the training pattern is added to the cluster of that category, 19
and the definition vector Ik of that category is modified to 20
incorporate the effects of the feature vector IINP of the input 21
pattern. If more than one category has a correlation CTRN above 22
the threshold λTRN, Ik for the best match category, i.e., the category 23
with the highest correlation, is modified and the training pattern is 24
added to the cluster of that category (Menon 5:38 – 6:7). 25
Wu 26
Appeal 2008-004750
Application 09/879,491

11
08. Wu is directed to providing Web product managers with quick 1
and detailed feedback on a visitor's satisfaction of the Web 2
product managers' own and competitive products. Specifically, 3
Wu aids a customer in designing a usability test for typical tasks 4
faced by a visitor to the customer's site. Wu administers a 5
usability test to a pre-qualified pool of testers meeting desired 6
demographic constraints. The usability tests measure a visitor's 7
success in achieving the visitor's objectives and also prompt for 8
context-specific feedback ranging from the aesthetics of the 9
design of the customer's site to a reason why a page request was 10
terminated. Statistics are aggregated across the testing population 11
and are presented as data with recommended actions backed up by 12
analysis (Wu 4:26-50). 13
09. Wu describes a sample test script for testing the usability of a 14
web site. This script states that among the implicit data to be 15
collected are links clicked on, links seen per page, and number of 16
distinct sites visited (Wu 18:Table B). 17
10. Wu describes combining collected data with data from other 18
clients in the analysis. The data sent may either be in raw form, or 19
summary statistics after processing has been performed at the 20
client (Wu 32:2-6). 21
11. Wu describes using its test to determine if a purchase rate 22
increase was due to better navigational cues and other factors (Wu 23
36: 24-31). 24
12. Wu describes performing additional analysis on the gathered 25
data. This analysis may include simple aggregation (sums and 26
Appeal 2008-004750
Application 09/879,491

12
averages, for example), selection (production of a subsample) of 1
"typical" data, finding outliers and either excluding them or 2
focusing on them, measuring correlations between data factors, 3
measuring the confidence in a hypothesis (Wu 33:48-55). 4
Facts Related To Differences Between The Claimed Subject Matter And The 5
Prior Art 6
13. None of the references describe determining if a training data 7
set or testing data set are geographically representative of a 8
customer population represented by the customer database. 9
14. None of the references describe obtaining data network 10
geographical information comprising frequency distributions of 11
the number of data network links between a customer 12
geographical location and one or more web site data network 13
geographical locations. 14
Facts Related To The Level Of Skill In The Art 15
15. Neither the Examiner nor the Appellant has addressed the level 16
of ordinary skill in the pertinent arts of systems analysis and 17
programming, predictive systems, training systems, and customer 18
analysis. We will therefore consider the cited prior art as 19
representative of the level of ordinary skill in the art. See Okajima 20
v. Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir. 2001) (“[T]he 21
absence of specific findings on the level of skill in the art does not 22
give rise to reversible error ‘where the prior art itself reflects an 23
appropriate level and a need for testimony is not shown’”) 24
(quoting Litton Indus. Prods., Inc. v. Solid State Sys. Corp., 755 25
F.2d 158, 163 (Fed. Cir. 1985)). 26
Facts Related To Secondary Considerations 27
Appeal 2008-004750
Application 09/879,491

13
16. There is no evidence on record of secondary considerations of 1
non-obviousness for our consideration. 2
3
PRINCIPLES OF LAW 4
Claim Construction 5
During examination of a patent application, pending claims are 6
given their broadest reasonable construction consistent with the 7
specification. In re Prater, 415 F.2d 1393, 1404-05 (CCPA 1969); In 8
re Am. Acad. of Sci. Tech Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004). 9
Limitations appearing in the specification but not recited in the claim are 10
not read into the claim. E-Pass Techs., Inc. v. 3Com Corp., 343 F.3d 1364, 11
1369 (Fed. Cir. 2003) (claims must be interpreted “in view of the 12
specification” without importing limitations from the specification into the 13
claims unnecessarily). 14
Although a patent applicant is entitled to be his or her own lexicographer 15
of patent claim terms, in ex parte prosecution it must be within limits. In re 16
Corr, 347 F.2d 578, 580 (CCPA 1965). The applicant must do so by placing 17
such definitions in the specification with sufficient clarity to provide a 18
person of ordinary skill in the art with clear and precise notice of the 19
meaning that is to be construed. See also In re Paulsen, 30 F.3d 1475, 1480 20
(Fed. Cir. 1994) (although an inventor is free to define the specific terms 21
used to describe the invention, this must be done with reasonable clarity, 22
deliberateness, and precision; where an inventor chooses to give terms 23
uncommon meanings, the inventor must set out any uncommon definition in 24
some manner within the patent disclosure so as to give one of ordinary skill 25
in the art notice of the change). 26
Appeal 2008-004750
Application 09/879,491

14
Statutory Subject Matter 1
[Whether a] patent is invalid for failure to claim statutory 2
subject matter under § 101, is a matter of both claim 3
construction and statutory construction. 4
State St. Bank & Trust Co. v. Signature Fin. Group, 149 F.3d 1368, 1370 5
(Fed. Cir. 1998). 6
Whoever invents or discovers any new and useful process, 7
machine, manufacture, or composition of matter, or any new 8
and useful improvement thereof, may obtain a patent therefor, 9
subject to the conditions and requirements of this title. 10
35 U.S.C. § 101. 11
[T]he Court has held that a claim is not a patent-eligible 12
“process” if it claims “laws of nature, natural phenomena, [or] 13
abstract ideas.” Diamond v. Diehr, 450 U.S. 175, 185, 101 14
S.Ct. 1048, 67 L.Ed.2d 155 (1981) (citing Flook, 437 U.S. at 15
589, 98 S.Ct. 2522, and Gottschalk v. Benson, 409 U.S. 63, 67, 16
93 S.Ct. 253, 34 L.Ed.2d 273 (1972)). Such fundamental 17
principles [as “laws of nature, natural phenomena, [or] abstract 18
ideas”] are “part of the storehouse of knowledge of all men ... 19
free to all men and reserved exclusively to none.” Funk Bros. 20
Seed Co. v. Kalo Inoculant Co., 333 U.S. 127, 130, 68 S.Ct. 21
440, 92 L.Ed. 588 (1948); see also Le Roy v. Tatham, 55 U.S. 22
(14 How.) 156, 175, 14 L.Ed. 367 (1852) (“A principle, in the 23
abstract, is a fundamental truth; an original cause; a motive; 24
these cannot be patented, as no one can claim in either of them 25
an exclusive right.”). “Phenomena of nature, though just 26
discovered, mental processes, and abstract intellectual concepts 27
are not patentable, as they are the basic tools of scientific and 28
technological work.” Benson, 409 U.S. at 67. . . . 29
In re Bilski, 545 F.3d 943, 952 (Fed. Cir. 2008) (footnote omitted). 30
The Court in Diehr thus drew a distinction between those 31
claims that “seek to pre-empt the use of” a fundamental 32
principle, on the one hand, and claims that seek only to 33
foreclose others from using a particular “application” of that 34
fundamental principle, on the other. 450 U.S. at 187, 101 S.Ct. 35
Appeal 2008-004750
Application 09/879,491

15
1048. Patents, by definition, grant the power to exclude others 1
from practicing that which the patent claims. Diehr can be 2
understood to suggest that whether a claim is drawn only to a 3
fundamental principle is essentially an inquiry into the scope of 4
that exclusion; i.e., whether the effect of allowing the claim 5
would be to allow the patentee to pre-empt substantially all uses 6
of that fundamental principle. If so, the claim is not drawn to 7
patent-eligible subject matter. 8
Id. 545 F.3d at 953. 9
The Supreme Court . . . has enunciated a definitive test to 10
determine whether a process claim is tailored narrowly enough 11
to encompass only a particular application of a fundamental 12
principle rather than to pre-empt the principle itself. A claimed 13
process is surely patent-eligible under § 101 if: (1) it is tied to a 14
particular machine or apparatus, or (2) it transforms a particular 15
article into a different state or thing. See Benson, 409 U.S. at 16
70, 93 S.Ct. 253 (“Transformation and reduction of an article 17
‘to a different state or thing’ is the clue to the patentability of a 18
process claim that does not include particular machines.”); 19
Diehr, 450 U.S. at 192, 101 S.Ct. 1048 (holding that use of 20
mathematical formula in process “transforming or reducing an 21
article to a different state or thing” constitutes patent-eligible 22
subject matter); see also Flook, 437 U.S. at 589 n.9, 98 S.Ct. 23
2522 (“An argument can be made [that the Supreme] Court has 24
only recognized a process as within the statutory definition 25
when it either was tied to a particular apparatus or operated to 26
change materials to a ‘different state or thing’ ”); Cochrane v. 27
Deener, 94 U.S. 780, 788, 24 L.Ed. 139 (1876) (“A process is 28
. . . an act, or a series of acts, performed upon the subject-29
matter to be transformed and reduced to a different state or 30
thing.”). 31
32
Id. 545 F.3d at 954 (footnote omitted). 33
The machine-or-transformation test is a two-branched inquiry; 34
an applicant may show that a process claim satisfies § 101 35
either by showing that his claim is tied to a particular machine, 36
Appeal 2008-004750
Application 09/879,491

16
or by showing that his claim transforms an article. See Benson, 1
409 U.S. at 70, 93 S.Ct. 253. Certain considerations are 2
applicable to analysis under either branch. First, as illustrated 3
by Benson and discussed below, the use of a specific machine 4
or transformation of an article must impose meaningful limits 5
on the claim's scope to impart patent-eligibility. See Benson, 6
409 U.S. at 71-72, 93 S.Ct. 253. Second, the involvement of 7
the machine or transformation in the claimed process must not 8
merely be insignificant extra-solution activity. See Flook, 437 9
U.S. at 590 . . . . 10
Id. 545 F.3d at 961-62. 11
Written Description 12
The first paragraph of 35 U.S.C. § 112 requires that the specification 13
shall contain a written description of the invention. This requirement is 14
separate and distinct from the enablement requirement. See, e.g., Vas-Cath, 15
Inc. v. Mahurkar, 935 F.2d 1555, 1563-64 (Fed. Cir. 1991). 16
The “written description” requirement implements the principle 17
that a patent must describe the technology that is sought to be 18
patented; the requirement serves both to satisfy the inventor's 19
obligation to disclose the technologic knowledge upon which 20
the patent is based, and to demonstrate that the patentee was in 21
possession of the invention that is claimed. 22
23
Capon v. Eshhar, 418 F.3d 1349, 1357 (Fed. Cir. 2005). 24
One shows that one is “in possession” of the invention by 25
describing the invention, with all its claimed limitations, not 26
that which makes it obvious. Id. (“[T]he applicant must also 27
convey to those skilled in the art that, as of the filing date 28
sought, he or she was in possession of the invention. The 29
invention is, for purposes of the ‘written description’ inquiry, 30
whatever is now claimed.”) (emphasis in original). One does 31
that by such descriptive means as words, structures, figures, 32
diagrams, formulas, etc., that fully set forth the claimed 33
invention. Although the exact terms need not be used in haec 34
verba, see Eiselstein v. Frank, 52 F.3d 1035, 1038 . . . (Fed. 35
Appeal 2008-004750
Application 09/879,491

17
Cir.1995) (“[T]he prior application need not describe the 1
claimed subject matter in exactly the same terms as used in the 2
claims . . . .”), the specification must contain an equivalent 3
description of the claimed subject matter. 4
5
Lockwood v. Am. Airlines, Inc., 107 F.3d 1565, 1572 (Fed. Cir. 1997). 6
It is the disclosures of the applications that count. Entitlement 7
to a filing date does not extend to subject matter which is not 8
disclosed, but would be obvious over what is expressly 9
disclosed. It extends only to that which is disclosed. While the 10
meaning of terms, phrases, or diagrams in a disclosure is to be 11
explained or interpreted from the vantage point of one skilled in 12
the art, all the limitations must appear in the specification. The 13
question is not whether a claimed invention is an obvious 14
variant of that which is disclosed in the specification. Rather, a 15
prior application itself must describe an invention, and do so in 16
sufficient detail that one skilled in the art can clearly conclude 17
that the inventor invented the claimed invention as of the filing 18
date sought. 19
Id. at 1571-72. 20
Obviousness 21
A claimed invention is unpatentable if the differences between it and 22
the prior art are “such that the subject matter as a whole would have been 23
obvious at the time the invention was made to a person having ordinary skill 24
in the art.” 35 U.S.C. § 103(a) (2000); KSR Int’l Co. v. Teleflex Inc., 550 25
U.S. 398, 406 (2007); Graham v. John Deere Co., 383 U.S. 1, 13-14 (1966). 26
In Graham, the Court held that that the obviousness analysis is 27
bottomed on several basic factual inquiries: “[(1)] the scope and content of 28
the prior art are to be determined; [(2)] differences between the prior art and 29
the claims at issue are to be ascertained; and [(3)] the level of ordinary skill 30
in the pertinent art resolved.” 383 U.S. at 17. See also KSR, 550 U.S. at 31
406-07. “The combination of familiar elements according to known 32
Appeal 2008-004750
Application 09/879,491

18
methods is likely to be obvious when it does no more than yield predictable 1
results.” Id. at 416. 2
“When a work is available in one field of endeavor, design incentives 3
and other market forces can prompt variations of it, either in the same field 4
or a different one. If a person of ordinary skill can implement a predictable 5
variation, § 103 likely bars its patentability.” Id. at 417. 6
“For the same reason, if a technique has been used to improve one 7
device, and a person of ordinary skill in the art would recognize that it would 8
improve similar devices in the same way, using the technique is obvious 9
unless its actual application is beyond his or her skill.” Id. 10
“Under the correct analysis, any need or problem known in the field 11
of endeavor at the time of invention and addressed by the patent can provide 12
a reason for combining the elements in the manner claimed.” Id. at 420. 13
14
ANALYSIS 15
Claims 1-8, 10-22, 24-35, and 37-43 rejected under 35 U.S.C. § 101 as 16
directed to non-statutory subject matter. 17
The Appellant argues each independent claim with the claims that 18
depend from it as a group. The sole exception is that claim 7 is separately 19
argued from parent claim 1. 20
The Examiner found that none of the claims recite a concrete and 21
tangible result. Although they recite using a predictive algorithm they do 22
not recite a concrete and tangible result from using the algorithm. The 23
Examiner also found that claims 29 and 43 do not meet the definition of a 24
true data structure. 25
Appeal 2008-004750
Application 09/879,491

19
The Appellant contends that the claims each fall within the 1
enumerated categories of statutory subject matter and produce non abstract 2
results (Br. 14-19). 3
With respect to method claim 1 and the claims depending therefrom, 4
we apply the machine-or-transformation test, as described in Bilski, to 5
determine whether the subject matter are patent-eligible under 35 U.S.C. § 6
101. 7
These claims recite a series of process steps that are not tied in any 8
manner to a machine. In other words, these claims do not limit the process 9
steps to any specific machine or apparatus. Thus, the claims fail the first 10
prong of the machine-or-transformation test because they are not tied to a 11
particular machine or apparatus. The steps of these process claims also fail 12
the second prong of the machine-or-transformation test because the data 13
does not represent physical and tangible objects.3 Rather, the data represents 14
information about a generic training and testing data set, which are 15
intangible data. Although the data is compared to a customer database, the 16
customer database is not transformed. Thus, the process of claim 1 and the 17
claims depending therefrom fails the machine-or-transformation test and is 18
not patent-eligible under 35 U.S.C. § 101. We note that the Appellant 19
separately argues claim 7 as generating recommendations (Br. 16). 20
Generating such recommendations transforms nothing. It merely creates 21
abstract subject matter, which is given no patentable weight. This claim 22
fails the machine-or-transformation test for the same reasons. 23

3 Because the data does not represent physical and tangible objects, we need
not reach the issue of whether mere calculation of a number based on inputs
Appeal 2008-004750
Application 09/879,491

20
Computer program product claim 29 and the claims dependent therefrom 1
recite instructions on a computer readable medium for executing the method 2
steps in claim 1 and its dependent claims. The issue presented by these 3
claims is whether recitation of such steps is more than the manipulation of 4
abstract ideas. We find that the steps performed by the instructions do no 5
more than generate arbitrary data sets, compare them, modify them, and then 6
use them in some unspecified predictive algorithm. Thus, since the data is 7
totally arbitrary and is no more than the abstract representation of ideas that 8
may be equally abstract, the computer program product contains instructions 9
that do no more than manipulate such abstract ideas. See In re Warmerdam, 10
33 F.3d 1354, 1360 (Fed. Cir. 1994). 11
Apparatus claim 15 does recite particular structural limitations such as 12
a statistical engine and comparison engine. Similarly, apparatus claim 42 13
recites means that are structurally identified in the Specification. Thus, we 14
find the apparatus claims are directed to specific machines and are 15
accordingly statutory subject matter. Process claim 41 and computer 16
program product claim 43 both recite training a machine and accordingly are 17
directed to machines that have such structure as may be adapted by training. 18
Therefore these claims are drawn to statutory subject matter as well. 19
The Appellant has not sustained its burden of showing that the 20
Examiner erred in rejecting claims 1-8, 10-14, 29-35, and 37-40 rejected 21
under 35 U.S.C. § 101 as directed to non-statutory subject matter. 22

of other numbers is a sufficient “transformation” of data to render a process
patent-eligible under § 101.
Appeal 2008-004750
Application 09/879,491

21
The Appellant sustained its burden of showing that the Examiner 1
erred in rejecting claims 15-22, 24-28, and 41-43 rejected under 35 U.S.C. § 2
101 as directed to non-statutory subject matter. 3
Claims 1, 15, 29, and 41-43 rejected under 35 U.S.C. § 112, first paragraph, 4
as lacking a supporting written description within the original disclosure. 5
The Examiner found that nowhere in the Appellant’s Specification is 6
it explained how the predictive algorithm would predict customer behavior 7
based upon network geographic location. 8
The Appellant contends that the Specification adequately describes 9
how to use the invention (Br. 19-23 and 23-27). Since the issue is the 10
description of use of the predictive algorithm, the Specification 45 11
describing the use of the algorithm is most pertinent (FF 03). The predictive 12
algorithm may be used to generate customer behavior predictions 13
(Specification 45:17-18). That is, the usage of the algorithm is simply 14
generating output from the disclosed algorithm. The Examiner found that 15
the Specification did not describe how to predict customer behavior, but the 16
claim only requires using the algorithm, not predicting behavior. The 17
Specification adequately describes such usage as generating prediction 18
output from the algorithm. 19
The Appellant has sustained its burden of showing that the Examiner 20
erred in rejecting claims 1, 15, 29, and 41-43 under 35 U.S.C. § 112, first 21
paragraph, as lacking a supporting written description within the original 22
disclosure. 23
Appeal 2008-004750
Application 09/879,491

22
Claims 1-8, 10-22, 24-35, and 37-43 rejected under 35 U.S.C. § 103(a) 1
as unpatentable over Menon, Wu, and Appellant’s Admitted Prior Art. 2
The Examiner found that Wu described determining if a training data 3
set or testing data set are geographically representative of a customer 4
population represented by the customer database (in claims 1-40) and 5
obtaining data network geographical information comprising frequency 6
distributions of the number of data network links between a customer 7
geographical location and one or more web site data network geographical 8
locations (in claims 41-43). 9
The Appellant contends that Wu does not describe such 10
determinations (Br. 28-30). The Examiner pointed to Wu’s Table B and 11
column 36, lines 24-30 (Answer 6) and column 24, lines 1-25 (Answer 12). 12
Table B refers to collecting links clicked on and pages visited. Wu 13
column 36, lines 24-30 states that its algorithm can discern whether a rise in 14
purchase rate is due to better navigational cues. Wu, column 24, lines 1-25 15
lists survey questions regarding site visits. 16
None of these makes reference to the number of links between a 17
customer geographical location and one or more web site data network 18
geographical locations or to determining if a training data set or testing data 19
set are geographically representative of a customer population represented 20
by the customer database. 21
The Examiner makes no attempt to map Wu to these specific 22
requirements, but only point us to table B, column 24, lines 1-25, and 23
column 36, lines 24-30. The Examiner does not say that Wu actually states 24
these limitations; only that it would be obvious to use these limitations with 25
Wu – but with no rationale. 26
Appeal 2008-004750
Application 09/879,491

23
Thus, the Examiner has not made findings as to why one of ordinary 1
skill would have included the claimed determining if at least one of the 2
training data set and the testing data set are geographically representative of 3
a customer population represented by the customer database, or obtaining 4
data comprising frequency distributions of number of data network links 5
between a customer geographical location and one or more web site data 6
network geographical locations using Wu, and has therefore failed to present 7
a prima facie case. 8
9
CONCLUSIONS OF LAW 10
The Appellant has not sustained its burden of showing that the 11
Examiner erred in rejecting claims 1-8, 10-14, 29-35, and 37-40 rejected 12
under 35 U.S.C. § 101 as directed to non-statutory subject matter. 13
The Appellant has sustained its burden of showing that the Examiner 14
erred in rejecting claims 15-22, 24-28, and 41-43 rejected under 35 U.S.C. § 15
101 as directed to non-statutory subject matter. 16
The Appellant has sustained its burden of showing that the Examiner 17
erred in rejecting claims 1, 15, 29, and 41-43 under 35 U.S.C. § 112, first 18
paragraph, as lacking a supporting written description within the original 19
disclosure. 20
The Appellant has sustained its burden of showing that the Examiner 21
erred in rejecting claims 1-8, 10-22, 24-35, and 37-43 under 35 U.S.C. § 22
103(a) as unpatentable over Menon, Wu, and Appellant’s Admitted Prior 23
Art. 24
Appeal 2008-004750
Application 09/879,491

24
DECISION 1
To summarize, our decision is as follows: 2
• The rejection of claims 1-8, 10-14, 29-35, and 37-40 rejected under 3
35 U.S.C. § 101 as directed to non-statutory subject matter is 4
sustained. 5
• The rejection of claims 15-22, 24-28, and 41-43 rejected under 6
35 U.S.C. § 101 as directed to non-statutory subject matter is not 7
sustained. 8
• The rejection of claims 1, 15, 29, and 41-43 under 35 U.S.C. § 112, 9
first paragraph, as lacking a supporting written description within the 10
original disclosure is not sustained. 11
• The rejection of claims 1-8, 10-22, 24-35, and 37-43 under 12
35 U.S.C. § 103(a) as unpatentable over Menon, Wu, and Appellant’s 13
Admitted Prior Art is not sustained. 14
No time period for taking any subsequent action in connection with 15
this appeal may be extended under 37 C.F.R. § 1.136(a)(1)(iv) (2007). 16
17
AFFIRMED-IN-PART 18
19
20
21
22
hh 23
Duke Yee 24
Yee & Associates P C 25
4100 Alpha Road Suite 1100 26
Dallas, TX 75244 27