Facebook, Inc.

Patent Trials and Appeals Board

Mar 2, 2021

2020004322 (P.T.A.B. Mar. 2, 2021)

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.
14/716,772 05/19/2015 Haibin Cheng 2006.043US1 7160
160546 7590 03/02/2021
Mannava & Kang, P. C.
3201 Jermantown Road
Suite 525
Fairfax, VA 22030
EXAMINER
STROUD, CHRISTOPHER
ART UNIT PAPER NUMBER
3688
NOTIFICATION DATE DELIVERY MODE
03/02/2021 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@mannavakang.com
fb-pdoc@fb.com
PTOL-90A (Rev. 04/07)
Appeal 2020-004322
Application 14/716,772
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________
Ex parte Haibin Cheng, Sue Ann Hong, Rituraj Kirti, Benjamin Tucker
Savage, and Gary Wu
Appeal 2020-004322
Application 14/716,772
Technology Center 3600
____________
Before JOSEPH A. FISCHETTI, BIBHU R. MOHANTY, and
CYNTHIA L. MURPHY, Administrative Patent Judges.
MURPHY, Administrative Patent Judge.
1 DECISION ON APPEAL
The Appellant appeals from the Examiner’s rejections of
claims 7–11, 13–19, and 21–26 under 35 U.S.C § 101 (Rejection I)
and 35 U.S.C. § 103 (Rejections II–IV). We sustain all of the
Examiner’s rejections, and, therefore, we AFFIRM.
Appeal 2020-004322
Application 14/716,772
2
2 THE APPELLANT’S INVENTION
The Appellant’s invention relates to a method for determining
“target audiences” for an advertising (“ad”) campaign. (Spec. ¶ 1.)
The Appellant’s method is practiced during an “exploration
period,” and, during this exploration period, the criteria used to
generate a “target audience” is iteratively refined. (Spec. ¶ 4.) The
Appellant describes this iterative refinement in terms of the target
audiences shown in Figure 5, below.
Figure 5 “shows an example of modifying a target audience using
seed groups of users during an exploration period for an ad
Appeal 2020-004322
Application 14/716,772
3
campaign.” (Id. ¶ 11.) In the above drawing, a first target
audience 510, a second target audience 530, a third target
audience 550, and a final target audience 570 each consists of ten
people. (See Spec. ¶¶ 60–62.)
With the Appellant’s method, there is a “performance objective”
for the ad campaign, a “threshold” determining whether a further
iteration is necessary, and an “end condition” for the exploration
period. (Spec. ¶ 59.) For example, a logical “performance objective”
of an advertiser is for people in a targeted audience to purchase the
project featured in its advertisement. (See id. ¶¶ 2, 53.) An “end
condition” could be when 50% of the targeted audience performs this
objective. (See id. ¶ 59.) Thus, when a target audience consists of
ten people, the end condition would be met when five people fulfill the
performance objective. (See id. ¶ 62.)
When the advertisement is presented to the initial target
audience 510, a subset 515 of two people (20%) fulfill the
performance objective. (Spec. ¶ 60.) Thus, the “end condition” of
50% performance is not met, the “exploration period continues,” and
the subset 515 is used to refine the targeting criteria. (Id.) For
example, if the two objective-performing people live in the same area,
the criteria is revised to target people living in corresponding areas.
(See id. ¶¶ 19, 34.) This revised targeting criteria “is used to
generate [the] second target audience 530.” (Id. ¶ 60.)
When the advertisement is presented to the second target
audience 530, a subset 535 of three people (30%) “fulfill the
performance objective.” (Spec. ¶ 61.) Thus, the “end condition” of
50% performance is not met, the “exploration period continues,” and
the subset 535 of the three object-performing people is used to
further refine the targeting criteria. (Id.) For example, if the three
objective-performing people fall within the same age range, the
Appeal 2020-004322
Application 14/716,772
4
criteria is further refined to target people falling within this age range.
(See id. ¶¶ 19, 34.) This further revised targeting criteria is “used to
generate [the] third target audience 550.” (Id. ¶ 61.)
When the advertisement is presented to the third target
audience 550, a subset 555 of five people (50%) “fulfill the
performance objective.” (Spec. ¶ 62.) Thus, the “end condition” of
50% performance is met, the “exploration period” is over, and “the
third target audience 550 becomes the final target audience 570” for
the remaining duration of the ad campaign. (Id.)
Thus, with the Appellant’s method, targeting criteria is iteratively
refined as more information about people’s positive responses to an
advertisement becomes available.
However, the Appellant’s iterative refinement of the target
audience does not continue in an endless loop ad infinitum, as it only
continues until the “end condition” is met. (See Spec. ¶ 48.) In the
above example, the “end condition” was a “performance goal” which
is “a threshold of users fulfilling the performance objective.” (Id.)
Another possible end condition “is when the performance of a
particular target audience does not reach a threshold improvement
over a prior target audience’s performance goal.” (Id.)
Thus, with the Appellant’s method, if the current targeting
criteria is yielding successful results, and further refinement of the
targeting criteria would be of little help, there is no reason to waste
resources by allowing the iterations to continue.
3 ILLUSTRATIVE CLAIM
7.A method comprising:
receiving a request for an ad campaign from an
advertiser, the request for the ad campaign
comprising an advertisement, targeting criteria
specifying a first set of user characteristics, and one
or more performance objectives specifying user
actions performed subsequent to receipt of the
Appeal 2020-004322
Application 14/716,772
5
advertisement;
for one or more exploration periods, iteratively repeating
steps comprising:
determining an initial target audience of users who
are eligible to be served the advertisement;
receiving performance data about the advertisement
presented to the initial target audience, the
performance data indicating which subset of
the initial target audience performed the user
actions defined by the one or more
performance objectives of the ad campaign;
responsive to a number of users in the subset of the
initial target audience that performed the user
actions being less than a threshold number,
training a user model using the subset of the initial
target audience, the user model trained using
a second set of user characteristics
associated with the subset of the initial target
audience different from the first set of user
characteristics;
identifying a group of users of the social networking
system who have user characteristics similar
to the second set of user characteristics using
the user model;
classifying the identified group of users who have
the user characteristics similar to the second
set of user characteristics as a revised target
audience for the ad campaign;
responsive to an end condition occurring,
determining that the revised target audience is
a final target audience; and
responsive to the end condition not occurring,
repeating the steps with the revised target
audience as the initial target audience; and
presenting the ad campaign to the final target audience
for a remaining duration of the ad campaign
following the one or more exploration periods.
4 REJECTION I
The Examiner rejects claims 7–11, 13–19, and 21–26 under
35 U.S.C. § 101 as directed to a judicial exception (i.e., an abstract
idea) without significantly more. (Final Action 2.) To wit, the
Examiner concludes that the claims on appeal do not satisfy the Alice
Appeal 2020-004322
Application 14/716,772
6
Framework for patent eligibility. We have carefully considered the
Appellant’s arguments regarding the wrongness of this conclusion
(see Appeal Br. 7–12; Reply Br. 3–13) but we are not swayed
thereby. Thus, we sustain this rejection.
The 2019 § 101 Guidance
The 2019 Revised Patent Subject Matter Eligibility Guidance
(“2019 § 101 Guidance”) provides us with specific steps for
discerning whether a claim satisfies the Alice Framework for patent
eligibility. (See 2019 Revised Patent Subject Matter Eligibility
Guidance, 84 Fed. Reg. 50 (Jan. 7, 2019).) These steps are “[i]n
accordance with judicial precedent” and consist of a twopronged
Step 2A and a Step 2B. (Id. at 52.)
In the first prong of Step 2A (Prong One), a determination is
made as to whether the claim “recites” an abstract idea. (2019 § 101
Guidance, 84 Fed. Reg. at 54.) The Guidance “extracts and
synthesizes key concepts identified by the courts as abstract ideas,”
and these concepts include “[c]ertain methods of organizing human
activity,” and, more particularly “fundamental economic practices” and
“commercial” interactions, and even more particularly, “advertising.”
(Id. at 52.) For example, “targeted advertising” is an abstract idea.
(In re Morsa, 809 Fed.Appx. 913, 917 (Fed. Cir. 2020).) “[T]ailoring
information” presented in an advertisement to a particular audience is
“an abstract, overly broad concept long-practiced in our society.”
(Intellectual Ventures I LLC v. Capital One Bank (USA), 792 F.3d
1363, 1370 (Fed. Cir. 2015).)
In the second prong of Step 2A (Prong Two), a determination is
made as to whether the claim contains additional elements that
“integrate” the abstract idea “into a practical application.” (See 2019
§ 101 Guidance, 84 Fed. Reg. at 51.) “Additional elements” are
“claim features, limitations, and/or steps that are recited in the claim
Appeal 2020-004322
Application 14/716,772
7
beyond the identified judicial exception.” (Id. at 55 n.24) “A claim
that integrates a judicial exception into a practical application will
apply, rely on, or use the judicial exception in a manner that imposes
a meaningful limit on the judicial exception.” (Id. at 55.) For example,
when the additional elements are computer components, and the
claim “merely includes instructions to implement an abstract idea on a
computer or merely uses a computer as tool to perform an abstract
idea,” this is not enough to integrate the abstract idea into a practical
application. (Id. at 54, 55.)
In Step 2B, a determination is made as to whether the
additional elements recited in the claim, individually or in combination,
amount to “significantly more” than the abstract idea itself. (2019
§ 101 Guidance, Federal Register Vol. 84, No. 4, 56.) If the
additional elements consist of a conventional arrangement of well-
understood, routine, conventional computer components, they will not
amount to significantly more, and the claim does not comply with the
Alice Framework for patent eligibility. (Id.)
Analysis
As for Prong One of Step 2A, we agree with the Examiner that
independent claim 7 recites a fundamental economic practice and/or
commercial interaction, namely a targeted advertising strategy. (See
Final Action 4.)
Independent claim 7 recites steps involving “an ad campaign
from an advertiser,” “an advertisement,” “targeting criteria,” “user
characteristics,” a “target audience,” “performance objectives,” and
“performance data.” (Appeal Br., Claims App.) An advertisement,
targeting criteria, user characteristics, and a target audience are the
hallmarks of a targeted-advertising strategy. And, as the objective of
advertising is to prompt people to perform a specific action after
receiving the advertisement (e.g., buy a product), an advertiser would
Appeal 2020-004322
Application 14/716,772
8
want to know who is actually doing so (i.e., “performance data”).
Independent claim 7 recites steps involving “subset[s]” of a
target audience that perform “user actions” aligned with the
advertiser’s “performance objectives,” “user characteristics”
associated with these subsets, “group[s] of users” who have user
characteristics “similar to” those associated with the subsets,’ and
“revised” target audiences aligned with these groups of users.
(Appeal Br., Claims App.) In other words, claim 7 recites iteratively
revising targeting criteria as more information becomes available
regarding users’ positive responses to the advertisement. As
indicated above, an advertiser would want to know who is positively
responding to its advertisement, and it follows that, upon obtaining
this knowledge, the advertiser would want the targeting criteria
revised accordingly.
Independent claim 7 recites that iterative revising of the
targeting criteria occurs during “one or more exploration periods”
during which the “number of users” in the subset of a target audience
is “less than a threshold number.” (Appeal Br., Claims App.) Claim 7
also recites that “presenting the ad campaign to the final target
audience for a remaining duration of the ad campaign following the
one or more exploration periods.” (Id.) In other words, the advertiser
decides when the positive response to its advertisement deems no
further revisions to targeting criteria necessary, and proceeds with the
current targeting criteria. Such a decision is embedded, either
literally or allegorically, in a targeted advertising strategy.
As for Prong Two of Step 2A, we agree with the Examiner that
the abstract idea recited in independent claim 7 “is not integrated into
a practical application.” (Final Action 4.)
Independent claim 7 recites “training a user model using the
subset of the initial target audience, the user model trained using a
Appeal 2020-004322
Application 14/716,772
9
second set of user characteristics associated with the subset of the
initial target audience different from the first set of user
characteristics” and “identifying a group of users of the social
networking system who have user characteristics similar to the
second set of user characteristics using the user model.” (Appeal Br.,
Claims App.) Thus, claim 7 merely requires the “model” to
generically apply the abstract-idea-embedded steps discussed above
in a social-networking-system environment. This generic recital of an
allegedly computer-related component, and/or this casual linking of
the claimed method to an allegedly technical environment, does not
render the claim any less abstract. (See 2019 § 101 Guidance, 84
Fed. Reg. at 55.)
As for Step 2B, we agree with the Examiner that, inasmuch as
independent claim 7 recites computer components, they do not
“amount to significantly more than the judicial exception.” (Final
Action 4.) The Specification describes any additional elements of
concern as conventional computer components arranged in a
conventional manner. (See Spec. ¶¶ 5–18, Figs. 1, 2.)
Consequently, we agree with the Examiner that independent
claim 7 does not pass muster under the Alice Framework for patent
eligibility.
The Appellant’s Arguments
The Appellant argues that “the claims do not recite the concept
of advertising itself.” (Appeal Br. 9; see also Reply Br. 6.) However,
as outlined above, independent claim 7 recites many features that are
hallmarks of, or embedded in, a targeting advertising strategy. Also,
as pointed out by the Examiner (see Answer 6), the Specification
itself states that “[t]his invention relates generally to determining
target audiences for advertising, and specifically to adapting a target
audience based on fulfillment of performance objectives.” (Spec.
Appeal 2020-004322
Application 14/716,772
10
¶ 1.) Furthermore, controlling case law establishes that adapting a
target audience based on a person’s location and his/her
characteristics for the purpose of promoting an advertisement resides
comfortably within the realm of abstract ideas. (See Intellectual
Ventures, 792 F.3d. at 1369–70.)
The Appellant advances arguments premised upon
independent claim 7 reciting limitations (i.e., a “threshold” and an
“end condition”) that save “computational resources, the advertiser’s
resources, and time.” (Appeal Br. 11; see also Reply Br. .)
According to the Appellant, without these limitations, there would be a
“risk” of “overfitting the target group” and a “waste” of “computational
resources in trying to make marginal improvements to the target
audience.” (Id. at 10.) However, this would be true of any iterative
refinement of a target audience (or anything else for that matter) that
continues ad infinitum. We note that claim 7 does not specify any
particulars for the threshold and/or the end condition which would
guard against such overfitting and/or resource wasting.
The Appellant advances arguments premised upon
independent claim 7 requiring a computer system “to automatically
update the target audience of an ad campaign for the advertiser.”
(Appeal Br. 10.) According to the Appellant, such an automatic
updating “removes the burden of identifying an effective targeting
audience from an advertiser, who typically use a human focus group,
because the advertiser using the claimed invention would no longer
need to identify a target audience for the ad campaign by manual
guessing or wasteful investment of resources in human-centric
feedback methods to predict which types of users with whom the ad
campaign would perform best.” (Reply Br. 7.) The Appellant is
essentially saying that a practical application is achieved by using a
computer as a tool to “automate” an advertising strategy in which
Appeal 2020-004322
Application 14/716,772
11
targeting criteria is updated as audience-response information
becomes available. However, “the improved speed or efficiency”
resulting from applying a targeted advertising strategy “on a
computer” is not enough to transform it “into a patent-eligible
invention.” (Intellectual Ventures 792 F.3d at 1367.)
The Appellant advances arguments premised upon
independent claim 7 reciting a method “in an online social network
system setting” and “audience selection in the context of an online
system.” (Appeal Br. 11; see also Reply Br. 6–7.) According to the
Appellant, the claimed method uses “a process that is unique to an
online social networking system.” (Appeal Br. 11.) But claim 7 only
recites that an identified group of users is “of” a social networking
system. (Appeal Br., Claims App.) To the extent that this is enough
to generally link the claimed advertising method to a social
networking system, “the mere application of an abstract idea of
organizing human activity in a particular field is not sufficient to
integrate the judicial exception into a practical application.” (See
2019 § 101 Guidance, 84 Fed. Reg. at 55 n.22). For example, linking
the “abstract idea of hedging” to “the energy market” did not “make
the concept patentable.” (Bilski v. Kappos 561 U.S. 593, 612 (2010).)
The Appellant advances arguments premised upon
independent claim 7 requiring “a machine learning model trained and
used by a computer” (Appeal Br. 8; see also Reply Br. 5) which is
“trained using specific criteria and a specific technique” (Appeal
Br. 11). However, claim 7 only requires “training” the model using “a
second set of user characteristics associated with the subset of the
initial target audience different from the first set of user characteristic”
and “using” the model when “identifying a group of users of the social
networking system who have user characteristics similar to the
second set of user characteristics using the user model.” (Appeal Br.,
Appeal 2020-004322
Application 14/716,772
12
Claims App.) As discussed above, this “training” data can be as
simple as two users (out of ten users) living in the same area and this
“identifying” can be as simple as identifying a group of users living in
corresponding areas.
Thus, after careful consideration of the Appellant’s arguments,
we still agree with the Examiner that independent claim 7 does not
pass muster under the Alice Framework for patent eligibility.
Summary
We sustain the Examiner’s rejection of independent claim 7
under 35 U.S.C. § 101. The Appellant does not argue the rest of the
claims on appeal separately (see Appeal Br. 7–12), and so they fall
therewith.
5 REJECTIONS II–IV
The Examiner rejects claims 7, 8, 10, 11, 14–16, 18, 19, and 21
under 35 U.S.C. § 103 as unpatentable over Gerace, Li, and Saurabh
(Final Action 5); the Examiner rejects claims 9, 17, and 23 under
35 U.S.C. § 103 as unpatentable over Gerace, Li, Saurabh, Shah,
and Maycotte (id. at 10); and the Examiner rejects claims 13 and 26
under 35 U.S.C. § 103 as unpatentable over Gerace, Li, Saurabh,
and Yasnovsky (id. at 11). To wit, the Examiner concludes that the
claims on appeal would have been obvious over the prior art. We
have carefully considered the Appellant’s arguments regarding the
wrongness of this conclusion (see Appeal Br. 13–16; Reply
Br. 13–19) but we are not persuaded thereby. Thus, we sustain
these rejections.
Obviousness
“Rejections on obviousness grounds cannot be sustained by
mere conclusory statements; instead, there must be some
articulated reasoning with some rational underpinning to support the
Appeal 2020-004322
Application 14/716,772
13
legal conclusion of obviousness.” (KSR Int’l Co. v. Teleflex Inc., 550
U.S. 398, 418 (2007).) However, an obviousness evaluator “need
not seek out precise teachings directed to the specific subject matter
of the challenged claim” as “the inferences and creative steps that a
person of ordinary skill in the art would employ” can be taken into
account.” (Id. at 419.) For example, “[t]he combination of familiar
elements according to known methods is likely to be obvious when it
does no more than yield predictable results.” (Id. at 416.)
Analysis
We agree with the Examiner that Gerace teaches an
advertising method comprising the request-receiving, determining,
performance-data receiving, training, identifying, and classifying steps
set forth in independent claim 7. (See Final Action 5–7.)
Gerace discloses a method that “provides automatic targeting
of audiences (target users) and self-tailoring of target profiles.”
(Gerace c2:59–60.) To this end, Gerace’s method “records history of
users viewing the advertisements,” “performs a regression analysis
on the record history of users viewing the ads,” and “refines profiles
of target users based on the regression analysis.” (Id. c2:45–53.)
“As such, over time, the advertisements become better targeted to
users having an interest in said information.” (Id. c2:55–57.)
We agree with the Examiner that Gerace discloses an
“exploration period[]” in which the determining, performance-data
receiving, training, identifying, and classifying steps are performed.
(See Final Action 5–6.) With Gerace’s method, “an ad package is
shown to general users,” and the number of “hits” reveals which
users have an interest in this information. (Gerace c18:40–42.)
Gerace “runs a regression” to “see what characteristics are important,
and who (type of user profile) the ad appeals to most.” (Id.
c18:43–45.) “[W]eighting information based on that regression” is
Appeal 2020-004322
Application 14/716,772
14
automatically entered “to create a targeted system,” and the
advertisement is run again “in front of this new targeted group.” (Id.
c18:45–49.)
We agree with the Examiner that, in Gerace, the “exploration
period” ends with the “last iteration,” the “final target audience” is
defined by this last iteration, and that the advertisement is presented
to this “final target audience” for the duration of the ad campaign.
(See Final Action 13.) Gerace discloses a program that “runs a
regression every 10,000 hits.” (Gerace c18:49.) The last 10,000-hits
regression would an “end condition,” the target audience generated
by this last regression would be the “final target audience,” and the
advertisement would be presented to this final target audience for the
remainder of the ad campaign.
Consequently, we agree with the Examiner that the method set
forth in independent claim 7 would have been obvious over the
combined teachings of Gerace, Li, and Saurabh.
The Appellant’s Arguments
The Appellant advances arguments premised upon Gerace
disclosing that “the regression for updating the target group occurs
until the ad campaign ends.” (Appeal Br. 14.) Gerace does
disclose that its “auto-targeting” program “runs a regression every
10,000 hits” and “[t]his continues until the Ad package is exhausted.”
(Gerace c18:49–53.) Thus, according to the Appellant, the “end
condition” is the ad package being exhausted.
However, with Gerace’s auto-targeting program, the “last
regression” could instead be considered the “end condition” and it
could occur prior to the exhaustion of the ad package. Gerace
discloses that an ad package can be priced by a “number of hits,” and
thus an advertiser could purchase a “number of hits” that is not a
multiple of 10,000, leaving hits remaining after the last regression.
Appeal 2020-004322
Application 14/716,772
15
(Gerace c12:38–39) If so, the “end condition” would be the number
of hits left in the ad package dropping below 10,000.
For example, if the advertiser purchases 25,000 hits, the ad
package would not be exhausted after the second 10,000-hits
regression because 5,000 hits would still be remaining, but no further
regression would be run. Thus, the target audience after the second
10,000-hits regression would be the “final target audience” to which
the advertisement is presented until the 5,000 more hits are
accumulated.
Moreover, as discussed above, independent claim 7 does not
specify what parameters of an “end condition” would guarantee that it
would occur prior to the exhaustion of the ad campaign. And claim 7
also leaves open the question of what happens if the end condition
does not occur prior to the end of the ad campaign.
Furthermore, along this same line, independent claim 7’s
training step is “responsive to a number of users in the subset of the
initial target audience that performed the user actions being less than
[the] threshold number.” (Appeal Br., Claims App.) If this threshold is
exceeded before the first iteration of the exploration period, claim 7
leaves open the question of whether the training step, and the
identifying and classifying steps contingent thereon, must be
performed. And without the classifying step, there is no “revised
target audience for the ad campaign” for designation as either the
“final target audience” or the updated “initial target audience.” (Id.)
In any event, we agree with the Examiner that the method set
forth independent claim 7 would have been obvious over the prior art
of record.
Summary
We sustain the Examiner’s rejection of independent claim 7
under 35 U.S.C. § 103. The Appellant does not argue the rest of the
Appeal 2020-004322
Application 14/716,772
16
claims on appeal separately (see Appeal Br. 16, Reply Br. 18–19),
and so they fall therewith.
CONCLUSION
Claims
Rejected
35 U.S.C.
§
Reference(s)/
Basis
Affirmed Reversed
7–11, 13–19,
21–26
101 Eligibility 7–11, 13–19,
21–26
7, 8, 10, 11,
14–16, 18,
19, 21
103 Gerace, Li,
Saurabh
7, 8, 10, 11,
14–16, 18,
19, 21
9, 17, 23 103 Gerace, Li,
Saurabh, Shah,
Maycotte
9, 17, 23
13, 26 103 Gerace, Li,
Saurabh,
Yasnovsky
13, 26
OVERALL
OUTCOME
7–11, 13–19,
21–26
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)
(2020).
6 AFFIRMED