VMware, Inc.

Patent Trials and Appeals Board

Dec 21, 2020

2020003854 (P.T.A.B. Dec. 21, 2020)

UNITED STATES PATENT AND TRADEMARK OFFICE
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
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
14/080,661 11/14/2013 Dani Matzlavi B344 1059
152606 7590 12/21/2020
Olympic Patent Works PLLC
4979 Admiral Street
Gig Harbor, WA 98332
EXAMINER
GOLDBERG, IVAN R
ART UNIT PAPER NUMBER
3619
MAIL DATE DELIVERY MODE
12/21/2020 PAPER
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.
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte DANI MATZLAVI and MEIDAD ETZ-HADAR
___________

Appeal 2020-003854
Application 14/080,661
Technology Center 3600
____________

Before JASON V. MORGAN, ERIC B. CHEN, and JEREMY J. CURCURI,
Administrative Patent Judges.

CHEN, Administrative Patent Judge.

DECISION ON APPEAL

Appeal 2020-003854
Application 14/080,661

2
STATEMENT OF THE CASE
Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the
Examiner’s decision to reject claims 1–24. We have jurisdiction under
35 U.S.C. § 6(b).
We AFFIRM.

CLAIMED SUBJECT MATTER
The claims are directed to aiding an enterprise in deciding whether to
execute an application entirely within a private cloud or a hybrid
combination of the private cloud and public cloud services. (Abstract.)
Claim 1, reproduced below, is illustrative of the claimed subject
matter, with disputed limitations in italics:
1. A system for aiding an enterprise in deciding to
execute an application in a private cloud or in a hybrid private
cloud and public cloud, the system comprising:
one or more processors;
one or more data-storage devices; and
machine-readable instructions stored in the data-storage
devices that when executed using the one or more processors
performs operations comprising:
receiving a set of quantitative parameters
associated with running the application using
computational services provided by a public cloud
service provider;
receiving a set of quantitative organizational
parameters associated with the enterprise;

1 We use the word “Appellant” to refer to “applicant” as defined in
37 C.F.R. § 1.42(a). Appellant identifies the real party in interest as
VMWARE, INC. (Appeal Br. 1.)
Appeal 2020-003854
Application 14/080,661

3
dividing each of the quantitative parameters by a
square root of a sum of squares of the quantitative
parameters to generate corresponding normalized
quantitative parameters;
dividing each of the quantitative organizational
parameters by a square root of a sum of squares of the
quantitative organizational parameters to generate
corresponding normalized quantitative organizational
parameters;
training a decision model based on historical
quantitative parameters and historical quantitative
organizational parameters;
using the decision model to compute a
recommendation that indicates exclusive use of a private
cloud or use of a hybrid private cloud and public cloud to
execute the application based on the normalized
quantitative parameters and the normalized quantitative
organizational parameters; and
executing the application in the private cloud or
the hybrid private cloud and public cloud based at least
in part on the recommendation.

REFERENCES
Name Reference Date
Ferris et al. US 2011/0295999 A1 Dec. 1, 2011
JIAWEI HAN ET AL., DATA MINING: CONCEPTS AND TECHNIQUES 228–231
(3rd ed. 2012).
Ake Edlund & Maarten Koopmans, Practical Cloud Evaluation from a
Nordic eScience User Perspective, VIRTUALIZATION TECHNOLOGIES IN
DISTRIBUTED COMPUTING 29–37 (2011).
David Bazell & Yuan Peng, A Comparison of Neural Network Algorithms
and Preprocessing Methods for Star-Galaxy Discrimination, THE
ASTROPHYSICAL JOURNAL SUPPLEMENT SERIES 47–55 (1998).

Appeal 2020-003854
Application 14/080,661

4
REJECTIONS
Claims 1–5, 7–11, 13–17, and 19–24 stand rejected under 35 U.S.C.
§ 103 as being unpatentable over Ferris, Han, and Bazell.2
Claims 6, 12, and 18 stand rejected under 35 U.S.C. § 103 as being
unpatentable over Ferris, Han, Bazell, and Edlund.3

OPINION
§ 103 Rejection—Ferris, Han, and Bazell
First, we are unpersuaded by Appellant’s arguments (Appeal Br. 10–
11; see also Reply Br. 2) that the combination of Ferris, Han, and Bazell
would not have rendered obvious independent claim 1, which includes the
limitation “receiving a set of quantitative parameters associated with running
the application using computational services provided by a public cloud
service provider.”
The Examiner found that deployment rules 334 of Ferris, including
requirements for software programs (e.g., cloud resources, processor
capacity, or bandwidth capacity), correspond to the limitation “receiving a
set of quantitative parameters associated with running the application using
computational services provided by a public cloud service provider.” (Final
Act. 5–6; see also Ans. 3–5.) We agree with the Examiner’s findings.
Ferris relates to cloud computing, in particular, “analyzing cloud
deployment solutions taking into account the relative importance of various

2 The Examiner inadvertently omitted some dependent claims from the
statement of the rejection.
3 Appellant does not present any arguments with respect to the rejection of
dependent claims 6, 12, and 18 under 35 U.S.C. §103. Thus, any such
arguments are deemed to be waived.
Appeal 2020-003854
Application 14/080,661

5
cloud resources.” (¶ 1.) Figure 3 of Ferris illustrates decision system 302,
which communicates with clouds 304 and 306, via networks 308. (¶ 42.)
Ferris explains that “decision system [302] can be configured to obtain cloud
resource usage data representing cloud service consumption by a specific
user” (¶ 16) and includes customization module 332 that maintains a set of
deployment rules 334 (¶ 62). Moreover, Ferris explains that “deployment
rules 334 can be configured to store rules, algorithms, decision logic, or best
practices for deploying resources in a computing cloud environment such as
the clouds 304 and 306.” (¶ 62.) In particular, Ferris explains that “a rule
can specify that deployment architecture must include cloud resources
sufficient to meet the peak requirement for each resource that is recorded in
utilization data 338” or “a rule can specify that if an application’s processor
capacity utilization over the course of a period of time is less than ten
percent, then the application is an excellent candidate for virtualization.”
(¶ 63.) For example, Ferris provides the following: “baseline requirement
for a user’s computation-intensive application(s) is 7.0 Gbytes of memory
capacity, processor capacity equivalent to two 2.0 GHz, Intel™ Core 2™
T7200 processors, 700 Gbytes of storage capacity, and 1 Gbit/s of network
throughput capacity.” (¶ 89.) Because Ferris explains that decision system
302 obtains cloud resource usage data and includes deployment rules 334,
such as determining if cloud resources meet the peak requirement or process
capability utilization, Ferris teaches the limitation “receiving a set of
quantitative parameters associated with running the application using
computational services provided by a public cloud service provider.”
Appellant argues that “the first step describes an operation in which a
set of quantitative parameters are received as input” (Appeal Br. 10) and
Appeal 2020-003854
Application 14/080,661

6
“[t]he term ‘quantitative parameters’ in claim 1 is referring numerical
quantities or values and the first step of claim 1 describes receiving these
quantitative parameters” (id. at 11 (emphasis omitted); see also Reply Br. 2).
However, Ferris explains that “the decision system [302] can be configured
to obtain cloud resource usage data representing cloud service consumption
by a specific user.” (¶ 16; see also ¶¶ 62–63.)
Thus, we agree with the Examiner that the combination of Ferris, Han,
and Bazell would have rendered obvious independent claim 1, which
includes the limitation “receiving a set of quantitative parameters associated
with running the application using computational services provided by a
public cloud service provider.”
Second, we are unpersuaded by Appellant’s arguments (Appeal
Br. 16–17; see also Reply Br. 2–3) that the combination of Ferris, Han, and
Bazell would not have rendered obvious independent claim 1, which
includes the limitation “receiving a set of quantitative organizational
parameters associated with the enterprise.”
The Examiner found that resource importance data 335 of Ferris,
which assigns weighting factors 9 on a scale of 1–10, correspond to the
limitation “receiving a set of quantitative organizational parameters
associated with the enterprise.” (Final Act. 6–7; see also Ans. 5–9.) We
agree with the Examiner’s findings.
Ferris explains that “the set of resource importance data 335 can
include weighting data for cloud resources needed by the applications or
processes associated with a user, where the weighting data associated with a
cloud resource indicates that resource’s relative importance among the set of
needed resources.” (¶ 64; see also id. ¶ 25.) In particular, Ferris explains
Appeal 2020-003854
Application 14/080,661

7
that “the resource importance data for this case indicates high importance for
processing capacity resources (e.g., a weighting factor of 9 on a scale of 1–
10, ten being the most important), low importance for storage capacity
resources (e.g., a weighting factor of 2), and average importance for memory
capacity resources and throughput capacity resources (e.g., a weighting
factor of 5).” (¶ 89.) Because Ferris explains that resource importance
data 335 is provided with a weighting factor on a scale of 1–10 for cloud
resources, Ferris teaches the limitation “receiving a set of quantitative
organizational parameters associated with the enterprise.”
Appellant argues that “the second step describes an operation in which
a set of quantitative organizational parameters associated with the enterprise
are received as input” and “none of the paragraphs the Examiner cites in
Ferris describes an operation in which a set of parameters are received and
that the set of parameters satisfy the limitations described in the second
step.” (Appeal Br. 16; see also Reply Br. 2–3.) Again, as discussed
previously, Ferris explains that “the decision system [302] can be configured
to obtain cloud resource usage data representing cloud service consumption
by a specific user.” (¶ 16; see also ¶¶ 62–63.)
Appellant further argues that “Ferris mentions parameters in
paragraph [0025], but the parameters refer to user instantiation requests” and
“[t]here is no teaching or suggestion that the quantitative parameters of
paragraph [0025] are . . . associated with how the user is organized.”
(Appeal Br. 17.) However, Appellant’s arguments are not commensurate in
scope with claim 1, because the claim does not require “associated with how
the user is organized.” As discussed previously, the limitation
“organizational parameters associated with the enterprise” is broad enough
Appeal 2020-003854
Application 14/080,661

8
to encompass resource importance data 335 of Ferris, which provides a
weighting factor for cloud resources.
Thus, we agree with the Examiner that the combination of Ferris, Han,
and Bazell would have rendered obvious independent claim 1, which
includes the limitation “receiving a set of quantitative organizational
parameters associated with the enterprise.”
Third, we are unpersuaded by Appellant’s arguments (Appeal Br. 19–
21; see also Reply Br. 4–5) that the combination of Ferris, Han, and Bazell
would not have rendered obvious independent claim 1, which includes the
limitations “dividing each of the quantitative parameters by a square root of
a sum of squares of the quantitative parameters to generate corresponding
normalized quantitative parameters” and “dividing each of the quantitative
organizational parameters by a square root of a sum of squares of the
quantitative organizational parameters to generate corresponding normalized
quantitative organizational parameters.”
The Examiner found that normalizing input data of Bazell by dividing
each component by the square root of the sum of squares of all components,
corresponds to the limitation “dividing each . . . by a square root of a sum of
squares . . . to generate corresponding normalized quantitative organizational
parameters.” (Final Act. 8.) The Examiner concluded that it would have
been obvious to combine Ferris and Bazell because:
The application of the known technique in Bazell would have
yielded no more than the predictable outcome which one of
ordinary skill would have expected to achieve, i.e. the ability to
utilize the particular kind of normalizing in Bazell for the data
of Ferris . . . by dividing each component by the square root of
the sum of squares of all the components and using this
normalized data.
Appeal 2020-003854
Application 14/080,661

9
(Final Act. 14.) We agree with the Examiner’s findings and conclusions.
Bazell relates to the “automatic processing of data.” (P. 47, col. 1,
§ 1.) In particular, Bazell explains that “[t]he input data were normalized by
dividing each component of the input vector by the square root of the sum of
squares of all components.” (P. 50, col. 1, § 3.2.)
Thus, the combination of Ferris and Bazell is nothing more than
applying the known data processing method of Bazell, in which input data
are normalized by dividing each component of the data by the square root of
the sum of squares, with the known deployment rules 334 and resource
importance data 335 of Ferris, to yield predictable results. See KSR Int’l Co.
v. Teleflex, Inc., 550 U.S. 398, 416 (2007) (“The combination of familiar
elements according to known methods is likely to be obvious when it does
no more than yield predictable results.”). This combination of Ferris and
Bazell would result in the processing deployment rules 334 and resource
importance data 335 of Ferris using the known normalization method of
Bazell, in which deployment rules 334 and resource importance data 335 are
divided by the square root of the sum of squares of the data. Accordingly,
the combination of Ferris and Bazell teaches the limitations “dividing each
of the quantitative parameters by a square root of a sum of squares of the
quantitative parameters to generate corresponding normalized quantitative
parameters” and “dividing each of the quantitative organizational parameters
by a square root of a sum of squares of the quantitative organizational
parameters to generate corresponding normalized quantitative organizational
parameters.” Thus, we agree with the Examiner (Final Act. 14) that
modifying Ferris with Bazell would have been obvious.
Appeal 2020-003854
Application 14/080,661

10
Appellant argues the following:
[T]he Examiner has not provided an explanation of how the
normalization process described in the cited passages of Bazell
would be applied to the cited passages of Ferris.
(Appeal Br. 19.)
By not providing an explanation of how Bazell’s teaching of
normalization is applied to the descriptions in the cited passages
in paragraphs [0025], [0062], [0063], [0064], and [0089] of
Ferris to obtain the third and fourth steps of claim 1, the
Examiner appears to done nothing more than search prior art
reference for common terms and limitations without regard to
whether or the teachings of the reference can actually be
combined with a reasonable expectation of success.
(Id. at 21; see also Reply Br. 4–5.) Contrary to Appellant’s arguments, as
discussed previously, the combination of Ferris with Bazell is based on the
combination of familiar elements. (Final Act. 14.) Thus, the Examiner has
provided sufficient articulated reasoning with some rational underpinning
for combining Ferris and Bazell to support a conclusion of obviousness.
Thus, we agree with the Examiner that the combination of Ferris, Han,
and Bazell would have rendered obvious independent claim 1, which
includes the limitations “dividing each of the quantitative parameters by a
square root of a sum of squares of the quantitative parameters to generate
corresponding normalized quantitative parameters” and “dividing each of
the quantitative organizational parameters by a square root of a sum of
squares of the quantitative organizational parameters to generate
corresponding normalized quantitative organizational parameters.”
Fourth, we are unpersuaded by Appellant’s arguments (Appeal
Br. 23–24; see also Reply Br. 8–11) that the combination of Ferris, Han, and
Bazell would not have rendered obvious independent claim 1, which
Appeal 2020-003854
Application 14/080,661

11
includes the limitation “training a decision model based on historical
quantitative parameters and historical quantitative organizational
parameters.”
The Examiner found that the data classification and decision tree of
Han corresponds to the limitation “training a decision model.” (Final
Act. 9.) The Examiner concluded that it would have been obvious to
combine Ferris and Han because:
Han improves upon Ferris by explicitly disclosing having
learning/training of a decision model such as a decision tree.
One of ordinary skill in the art would be motivated to further
include training data and utilizing information gain to select the
best choice for efficiently traversing a decision system of Ferris
to use information gain to improve the decision model tree.
(Id. at 13 (citations omitted).) We agree with the Examiner’s findings and
conclusions.
As discussed previously, Figure 3 of Ferris illustrates decision system
302 which communicates with clouds 304 and 306, via networks 308 (¶ 42)
and explains that “deployment rules 334 can be configured to store rules,
algorithms, decision logic, or best practices for deploying resources in a
computing cloud environment such as the clouds 304 and 306” (¶ 62).
Han explains the following:
Decision tree induction is the learning of decision trees from
class-labeled training tuples. A decision tree is a flowchart-like
tree structure, where each internal node (nonleaf node) denotes
a test on an attribute, each branch represents an outcome of the
test, and each leaf node (or terminal node) holds a class label.
(P. 330, § 8.2 (emphasis omitted).)
Decision trees can handle multidimensional data. Their
representation of acquired knowledge in tree form is intuitive
and generally easy to assimilate by humans. The learning and
Appeal 2020-003854
Application 14/080,661

12
classification steps of decision tree induction are simple and
fast. In general, decision tree classifiers have good accuracy.
However, successful use may depend on the data at hand.
Decision tree induction algorithms have been used for
classification in many application areas such as medicine,
manufacturing and production, financial analysis, astronomy,
and molecular biology.
(Id. at 331.)
A person of ordinary skill in the art would have recognized that
incorporating the decision tree of Han, with decision system 302 of Ferris,
would improve Ferris by providing the ability to handle multidimensional
data, with simple and fast learning and classification steps, and good
accuracy. See KSR, 550 U.S. at 417 (“[I]f a technique has been used to
improve one device, and a person of ordinary skill in the art would recognize
that it would improve similar devices in the same way, using the technique is
obvious unless its actual application is beyond his or her skill.”).
Alternatively, combining Ferris and Han is nothing more than incorporating
the known decision tree of Han, with decision system 302 of Ferris, to yield
predictable results. See id. at 416 (“The combination of familiar elements
according to known methods is likely to be obvious when it does no more
than yield predictable results.”). This combination of Ferris and Hall would
result in applying the decision tree of Han to deployment rules 334 and
resource importance data 335 of Ferris for deploying computing resources,
and accordingly, teaches the limitation “training a decision model based on
historical quantitative parameters and historical quantitative organizational
parameters.” Thus, we agree with the Examiner (Final Act. 13) that
modifying Ferris to include the decision tree induction of Han would have
been obvious.
Appeal 2020-003854
Application 14/080,661

13
Appellant argues the following:
[H]ow exactly the decision tree described by Han is trained by
the information provided in Ferris is not explained by the
Examiner. Ferris does not teach or suggest use of a decision
tree. Ferris does not teach or suggest use of neural network. So
how exactly would one skilled in the art modify Ferris with the
decision tree of Han is unknown.
(Appeal Br. 23; see also Reply Br. 8–11.) Contrary to Appellant’s
arguments, as discussed previously, the combination of Ferris and Hall is
based on the improvement of a similar device in the same way as in the prior
art. (Final Act. 13.) Thus, the Examiner has provided sufficient articulated
reasoning with some rational underpinning for combining Ferris and Han to
support a conclusion of obviousness.
Appellant further argues that
Han does not teach or suggest using historical quantitative
parameters and historical quantitative organizational parameters
or the equivalents of such parameters to train a decision tree.
Ferris also does not teach or suggest using historical
quantitative parameters and the historical quantitative
organizational parameters to train a decision tree.
(Appeal Br. 23–24 (emphasis omitted).) However, the Examiner cited to
Ferris for teaching the limitations “quantitative parameters” (Final Act. 5–6)
and “quantitative organizational (id. at 6–7) parameters,” and cited Han for
teaching the limitation “training a decision model” (id. at 9). The rejection
of claim 1 is based on the combination of Ferris and Han, and Appellant
cannot show non-obviousness by attacking the references individually. See
In re Keller, 642 F.2d 413, 426 (CCPA 1981).
Thus, we agree with the Examiner that the combination of Ferris, Han,
and Bazell would have rendered obvious independent claim 1, which
includes the limitation “training a decision model based on historical
Appeal 2020-003854
Application 14/080,661

14
quantitative parameters and historical quantitative organizational
parameters.”
Last, we are unpersuaded by Appellant’s arguments (Appeal Br. 26)
that the combination of Ferris, Han, and Bazell would not have rendered
obvious independent claim 1, which includes the limitation “executing the
application in the private cloud or the hybrid private cloud and public cloud
based at least in part on the recommendation.”
The Examiner found that decision system 302 of Ferris, which can
generate customized deployment architectures, corresponds to the limitation
“executing the application in the private cloud or the hybrid private cloud
and public cloud based at least in part on the recommendation.” (Final
Act. 11.) We agree with the Examiner’s findings.
Ferris explains that:
decision system 302 . . . generate[s] customized deployment
architectures and to account for the relative importance of cloud
resources, such as an identifier of a running application, a start
date and time, a requester identifier, and data parameters such
as a configuration of a current deployment architecture (e.g., a
number of machines), a time duration, a number of time
intervals, a processor utilization, a network traffic level, a
storage utilization, a software license information.
(¶ 66.) Moreover, Ferris explains that decision system 302 “advise[s] the
user of resource usage information, deployment architecture options” and
“[d]eployment options can include computing resources, such as, for
example, one or more physical machines, one or more virtual machines, one
or more public clouds, one or more private clouds, or any combination
thereof.” (¶ 70.) Because Ferris explains that decision system 302 generates
customized deployment architectures options, for example, the combination
of public and private clouds, Ferris teaches the limitation “executing the
Appeal 2020-003854
Application 14/080,661

15
application in the private cloud or the hybrid private cloud and public cloud
based at least in part on the recommendation.”
Appellant argues the following:
Ferris does not teach the extra step of executing the applications
for which recommendations or deployment architecture have
been generated. Ferris teaches generating the recommendation
or deployment architectures only. . . . Paragraph [0092]
describes “assigning the determined deployment architecture to
the user's application(s)/process(es) such that they run on the
determined deployment architecture.”
(Appeal Br. 26.) However, the Examiner also cited to paragraphs 66 and 70
of Ferris for teaching the limitation “executing the application in the private
cloud or the hybrid private cloud and public cloud based at least in part on
the recommendation,” and Appellant has not provided any persuasive
evidence and arguments as to why the Examiner’s findings are in error.
Thus, we agree with the Examiner that the combination of Ferris, Han,
and Bazell would have rendered obvious independent claim 1, which
includes the limitation “executing the application in the private cloud or the
hybrid private cloud and public cloud based at least in part on the
recommendation.”
Accordingly, we sustain the rejection of independent claim 1 under 35
U.S.C. § 103. Claims 2–5 depend from claim 1, and Appellant has not
presented any additional substantive arguments with respect to these claims.
Therefore, we sustain the rejection of claims 2–5 under 35 U.S.C. § 103 for
the same reasons discussed with respect to independent claim 1.
Independent claims 7 and 13 recite limitations similar to those
discussed with respect to independent claim 1, and Appellant has not
presented any additional substantive arguments with respect to these claims.
Appeal 2020-003854
Application 14/080,661

16
We sustain the rejection of claims 7 and 13, as well as dependent claims 8–
11, 14–17, and 19–24 for the same reasons discussed with respect to claim 1.

CONCLUSION
We affirm the Examiner’s decision rejecting claims 1–24 under 35
U.S.C. § 103.
Because we have affirmed at least one ground of rejection with
respect to each claim on appeal, the Examiner’s decision is affirmed. See
37 C.F.R. § 41.50(a)(1).

DECISION
In summary:
Claims
Rejected
35
U.S.C. §
References Affirmed Reversed
1–5, 7–11,
13–17, 19–24
103 Ferris, Han,
Bazell
1–5, 7–11,
13–17, 19–24

6, 12, 18 103 Ferris, Han,
Bazell, Edlund
6, 12, 18
Overall
Outcome
1–24

TIME PERIOD FOR RESPONSE
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)(iv).

AFFIRMED