VIACCESS

Patent Trials and Appeals Board

Jul 24, 2020

14911079 - (D) (P.T.A.B. Jul. 24, 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.
14/911,079 02/09/2016 Mathieu Boivin 3C105-017US1 9599
69713 7590 07/24/2020
OCCHIUTI & ROHLICEK LLP
50 Congress Street
Suite 1000
Boston, MA 02109
EXAMINER
ULLAH, SHARIF E
ART UNIT PAPER NUMBER
2495
NOTIFICATION DATE DELIVERY MODE
07/24/2020 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):
INFO@ORPATENT.COM
PTOL-90A (Rev. 04/07)

UNITED STATES PATENT AND TRADEMARK OFFICE
________________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
________________

Ex parte MATHIEU BOIVIN
________________

Appeal 2019-004083
Application 14/911,079
Technology Center 2400
________________

Before JASON V. MORGAN, IRVIN E. BRANCH, and
DAVID J. CUTITTA II, Administrative Patent Judges.

MORGAN, Administrative Patent Judge.

DECISION ON APPEAL
STATEMENT OF THE CASE
Introduction
Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the
Examiner’s decision to reject claims 8–11 and 13–18. Claims 1–7 are
canceled. The Examiner objects to claim 12 “as being dependent upon a
rejected base claim, but [the Examiner finds claim 12] would be allowable if
rewritten in independent form including all of the limitations of the base

1 We use the word “Appellant” to refer to “applicant” as defined in 37
C.F.R. § 1.42. Appellant identifies the real party in interest as VIACCESS.
Appeal Br. 1 (Sept. 7, 2018) (all references to “Appeal Br.” are to this
original filing).
Appeal 2019-004083
Application 14/911,079

2

claim and any intervening claims.” Final Act. 19. An oral hearing was held
June 18, 2020. A transcript was entered into the record July 8, 2020. We
have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.
Summary of the disclosure
Appellant’s claimed subject matter relates to “a method for providing
a licence in a system for providing multimedia contents.” Spec.2 ¶ 2.
Representative claims (key limitations emphasized)
8. In a system for providing multimedia contents, a method
comprising the non-abstract computation of a license that
allows a terminal of a user of the system to access a multimedia
content, the license being dependent both on an identifier of the
multimedia content and on an identifier of the terminal or of the
user, wherein the non-abstract computation of the license
comprises the non-abstract steps of:
a) at a license server, receiving a request including the identifier
of the multimedia content and the identifier of the terminal or of
the user,
b) at the server, computing a license including at least one
access data needed by the terminal to access the multimedia
content, the access data being dependent on the identifier of the
multimedia content and on the identifier of the terminal or of
the user, and
c) at the server, transmitting, in response to the request, the
license computed in step b),
wherein computing the license further comprises, before step a),
executing a step e) that comprises
obtaining a first recommendation list for the user, the
first recommendation list containing a limited number of

2 Herein we cite to the amended version of the Specification, filed February
9, 2016.
Appeal 2019-004083
Application 14/911,079

3

identifiers of multimedia contents from among a larger
set of multimedia contents available in the system, the
multimedia contents identified by the first
recommendation list having a higher probability of being
accessed by the terminal of the user than the other
multimedia contents of the larger set, the first
recommendation list being constructed according to the
identifier of the user and a database of pre-stored data on
the basis of which it is possible to compute a probability
that any multimedia content available in the system will
be accessed by the terminal of the user,
automatically selecting, from the first recommendation
list, at least one multimedia content identifier, and,
for each automatically selected multimedia identifier,
pre-computing at least one access data using the
multimedia content identifier and the identifier of the
terminal or of the user, and
wherein, in step b), the server searches for the identifier of the
multimedia content received in step a) from among the
identifiers automatically selected in step e), then, in the event of
success, selects the pre-computed access data corresponding to
the identifier, and, in the event of failure, generates the access
data.
11. The method according to claim 8, further comprising
computing the workload of the server as a function of the
number of requests to be processed by the server over a time
interval, and of the duration of the time interval, and inhibiting
or triggering of step e), depending on the computed workload.
17. The method according to claim 8, further comprising
triggering step e) in response to receiving a list of identifiers of
users unsubscribed from the system.
The Examiner’s rejections and cited references
The Examiner rejects claims 8–10 and 13–16 under 35 U.S.C. § 103
as being unpatentable over Inoue et al. (US 2003/0005135 A1; published
Appeal 2019-004083
Application 14/911,079

4

Jan. 2, 2003) (“Inoue”) and Horvitz et al. (US 2005/0193414 A1; published
Sept. 1, 2005) (“Horvitz”). Final Act. 11–16.
The Examiner rejects claim 11 under 35 U.S.C. § 103 as being
unpatentable over Inoue, Horvitz, and Leighton et al. (US 2013/0013748 A1;
published Jan. 10, 2013) (“Leighton”). Final Act. 16–17.
The Examiner rejects claims 17–18 under 35 U.S.C. § 103 as being
unpatentable over Inoue, Horvitz, and Relan et al. (US 2005/0235307 A1;
published Oct. 20, 2005) (“Relan”). Final Act. 17–19.
ANALYSIS
Claims 8–10 and 13–16
The Examiner finds that Horvitz, by ranking selections “that the user
is most likely to be interested in viewing” (Horvitz ¶ 61), teaches or
suggests:
obtaining a first recommendation list for the user, the first
recommendation list containing a limited number of identifiers
of multimedia contents from among a larger set of multimedia
contents available in the system, the multimedia contents
identified by the first recommendation list having a higher
probability of being accessed by the terminal of the user than
the other multimedia contents of the larger set,
as recited in claim 8 (see Final Act. 13 (citing Horvitz ¶¶ 9, 61, 71–72); Ans.
5 (citing Horvitz ¶¶ 28–34, 40–45)).
Appellant contends the Examiner erred because Horvitz teaches a
“sorting value [that] is not a measure of the probability of [a selection] being
accessed by the terminal of the user.” Appeal Br. 4. Specifically, Appellant
argues that Horvitz teaches a “‘sorting value’ [that] is just the probability of
viewing the title divided by the size of that title.” Id. at 5 (citing Horvitz
Appeal 2019-004083
Application 14/911,079

5

¶ 71). Because a large-sized selection with a high probability of being
viewed could have a lower sorting value than a small-sized selection with a
low probability, Appellant argues the “sorting values do not reflect the
probability of being accessed.” Id.; see also Reply Br. 2–4.
Appellant’s arguments are unpersuasive because Appellant focuses
exclusively on the Horvitz use of a sorting value in its cache retention policy
(i.e., in determining which selections to keep and which ones to remove in
whole or in part). See Horvitz ¶ 71 (cited in Appeal Br. 5) (describing “a
cache retention policy” in which “a sorting value is determined for each title
by determining the probability that a system user would like to view the
information divided by the size of the information”). But the Examiner also
cites to the user interface in Horvitz “that allows a user to review available
selections quickly without unduly searching” by providing the user with a
“plurality of recommendations . . . ranked according to selections that the
user is most likely to be interested in viewing.” Horvitz ¶ 61 (cited in Final
Act. 13). Unlike the “sorting value” used in the cache retention policy of
Horvitz, this ranking of selections does not divide the probability of a
selection being viewed by the size of the selection. Therefore, we agree with
the Examiner that the Horvitz ranking of selections—as is used in providing
the user with a plurality of recommendations—teaches or suggests the
disputed obtaining a first recommendation list recitation of claim 8. See
Final Act. 13.
The Examiner also relies on Horvitz—which provides a user interface
that allows a user to “give the system a go ahead for automated policies for
downloading future television content” (Horvitz ¶ 44)—to teach or suggest
“automatically selecting, from the first recommendation list, at least one
Appeal 2019-004083
Application 14/911,079

6

multimedia content identifier,” as recited in claim 8 (Final Act. 13 (citing
Horvitz ¶ 44); see also Ans. 7–8).
Appellant contends the Examiner erred because “[t]here is no human
intervention” in the automatic selection step. Appeal Br. 9. Specifically,
Appellant argues the Examiner’s “use of ‘automatic’ would make it
impossible to distinguish between an automatic step and a non-automatic
step.” Reply Br. 9.
Appellant’s arguments are unpersuasive because even if a user
manually enables policies for downloading future television content in
Horvitz, execution of the policies is nonetheless automated. Horvitz ¶ 44. As
the Examiner correctly points out, Appellant’s comparison of the disputed
automatically selecting recitation to technologies such as automatic
transmissions or automated teller machines neglects the human element in
the use of these automated technology analogues. See Ans. 7–8; Appeal
Br. 9.
Appellant argues that Horvitz nonetheless fails to teach or suggest the
disputed automatically selecting recitation because Horvitz “discloses a
human who selects from a list.” Reply Br. 9. Appellant argues this “is not
‘automatically selecting’ because the act of ‘selecting’ is not being carried
out without human intervention.” Id. But Horvitz teaches that its user
interface both enables a user “to select the most desirable shows for later
automated caching of content” and enables a user to “give the system a go
ahead for automated policies for downloading future television content.”
Horvitz ¶ 44. That is, Horvitz has an embodiment where a human makes
selections from a list (i.e., where the selection is manual, not automatic) and
an embodiment where a human instead allows the system to use automated
Appeal 2019-004083
Application 14/911,079

7

policies (i.e., where the system automatically selects content rather than rely
on manual selection of content). Therefore, we agree with the Examiner that
the Horvitz user interface for enabling automated policies for downloading
future television content teaches or suggests the automatically selecting
recitation of claim 8. See Final Act. 13.
The Examiner also relies on Horvitz—which teaches providing the
recommendation list to the local cache system and delivering programs from
a remote program source system (Horvitz ¶ 45)—to teach or suggest that, in
the recited step for computing a license, “the server searches for the
identifier of the multimedia content” received in the step of receiving a
request “from among the identifiers automatically selected . . . then, in the
event of success, selects the pre-computed access data corresponding to the
identifier” (Ans. 10–11 (citing Horvitz ¶ 45); see also Final Act. 13–14
(citing Horvitz ¶¶ 71–73)).
Appellant contends the Examiner erred. Although Appellant agrees
that in Horvitz “[t]he user causes the content identifier to be sent to the
server” and “[t]he server sends the requested content to the user,” Appellant
argues that instead of the server searching for the identifier, the server user
would instead use “the content identifier to search for the content.” See
Reply Br. 12–13; see also Appeal Br. 14 (“The server now knows the
identifier of this content. There is obviously no reason to search for it. It is
right there.”).
Appellant’s arguments are unpersuasive because, as the Examiner
correctly finds, it would have been obvious for the server to search “for the
identifier of the multimedia content” before sending “the user the requested
content.” Ans. 10. That is, it would have been obvious to an artisan of
Appeal 2019-004083
Application 14/911,079

8

ordinary skill to label multimedia content available to be sent on request
with an identifier (i.e., a content identifier) so that the correct multimedia
content identified in the request could be sent. Moreover, the behavior taught
or suggested by Horvitz is consistent with the server searching recitation, as
evidenced by the Specification’s disclosure that “server 60 searches the
database 69 to find out whether the identifier 18 is associated with the
identifier of the multimedia content contained in the received request.” Spec.
¶ 72. That is, a fair reading of the disputed recitation in light of the
Specification is that the process entails looking for the identifier stored on
the server that matches (i.e., is associated with) the identifier in the request.
Appellant acknowledges that if the server has the requested content
(i.e., if the server’s search for the content identifier was successful), then
“[t]he server sends the requested content to the user.” See Reply Br. 12–13;
see also Ans. 10–11 (quoting Horvitz ¶ 45). Although Appellant contends
there is a patentable distinction between “content and the access data needed
to watch content” (Appeal Br. 3; see also id. at 16), the Examiner properly
relies on Inoue’s issuing of a license ticket, “information for allowing a user
to use . . . content” (Inoue ¶ 57), to teach or suggest sending not just content,
but also access data to allow for use of the content (see Final Act. 11–12
(citing Inoue ¶¶ 57–59)). Moreover, the sorting value of Horvitz, which “is
determined for each title by determining the probability that a system user
would like to view the information divided by the size of the information”
(Horvitz ¶ 71) represents another form of access data because this
information determines whether the content will or will not be stored in or
retained in local cache, and thus whether the content will be available for use
Appeal 2019-004083
Application 14/911,079

9

(id. (“titles are stored in the local cache from the highest sorting value to the
lowest sorting value until the local cache is full”)).
For these reasons, we agree with the Examiner that the combination of
Inoue and Horvitz teaches or suggests the claimed step of a server searching
for, and successfully finding, an identifier of the multimedia content. See
Final Act. 11–12; Ans. 10–11.
The Examiner relies on the Horvitz “methodology for storing
information in a local cache utilizing the value density bin packing method
and a cache retention policy” (Horvitz ¶ 71) to teach or suggest the claimed
server behavior in the event of a failure in searching for the identifier of the
multimedia content (Final Act. 13–14 (citing Horvitz ¶¶ 71–73); Ans. 6–7).
As discussed above, the sorting values used for storing information in a local
cache represent a form of access data.
Appellant argues the Examiner nonetheless erred because Horvitz
fails to teach or suggest generating the access data, in the event of failure in
the server’s search for the identifier of the multimedia content. Appeal Br. 7–
8. Appellant argues that “if a title did not even exist” in Horvitz, “there
would be no purpose to generating access data.” Id. at 7. Appellant also
argues the Examiner erroneously relied on teachings in Horvitz related to
actions that take “place after new titles have been received.” Reply Br. 7
(emphasis added); see also Appeal Br. 8.
We note that claim 8 is a method claim (even though it is a method
performed in a system) and, therefore, the disputed conditional recitation is
subject to the pertinent holding of Ex parte Schulhauser, Appeal 2013-
007847, 2016 WL 6277792, at *7–8 (Apr. 28, 2016) (precedential). That is,
because the scenario in which the identifier of the multimedia content is
Appeal 2019-004083
Application 14/911,079

10

successfully found represents a reasonably broad scenario, the Examiner
need not show that the prior art teaches or suggests the scenario where the
search for the identifier of the multimedia content fails. That is, in the event
of success in the search for the identifier of the multimedia content, the
generating of the access data need not occur. See also Spec. Fig. 2, ¶¶ 72, 75
(detailing conditional operations 142 (generating access data) and 146
(selected pre-computed access data)). But even if a conditional claim
recitation of a method claim need not be performed in the broadest
reasonable scenario, the prior art can still teach or suggest the conditional
claim recitation. Thus, even though unnecessary in this case, we analyze
Appellant’s arguments and the Examiner’s findings with respect to the
disputed conditional recitation.
Appellant’s arguments that the relevant operations in Horvitz would
not take place if the title exists (Appeal Br. 7) and are performed after new
titles have been received (Reply Br. 7) are unpersuasive because Horvitz
specifically teaches that inference system 18, which is “used to infer the
likelihood of a user’s preferences for content” (Horvitz ¶ 30), can “reside . . .
remotely at a remote server” (id.). Moreover, it would have been obvious
that the sorting value, being based in part on “the probability that a system
user would like to view the information” (id. at 71 (emphasis added)), would
not need to be calculated until the title was ready to be stored in the local
cache of the user (id. (quoted in Ans. 6–7)). Thus, Horvitz teaches or
suggests not just searching for the identifier of the multimedia content to
determine that the content is available, but also searching for the identifier of
the multimedia content to determine if a sorting value for the content can be
provided to enable storage of the content and, if this latter search fails (i.e., a
Appeal 2019-004083
Application 14/911,079

11

sorting value is not available), calculating the sorting value (i.e., generating
the access data).
For these reasons, even though the disputed conditional recitation
need not be performed given the broadest reasonable scenario, we agree with
the Examiner that the combination of Inoue and Horvitz teaches or suggests
the claimed step of a server searching for and failing to find an identifier of
the multimedia content, and generating access data in response thereto. See
Final Act. 11–12; Ans. 10–11.
Appellant also argues the Examiner erred in relying on the combined
teachings and suggestions of Inoue and Horvitz because “[i]t appears that the
proposed modification would result in providing fewer than all available
selections.” Appeal Br. 9; compare with Final Act. 14 (citing Horvitz ¶ 9)
(“The suggestion/motivation [for] the combination would have been to
provide a subset of available viewing selections by assigning a value or
utility to a selection and retaining selections in the cache depending on the
value and size of a selection.”); see also Ans. 8–9. That is, Appellant argues
“there is no obvious reason to provide only a subset of what is available. The
obvious thing to do is to provide all that is available, not less than what is
available.” Appeal Br. 10; see also Reply Br. 10–12.
Appellant’s arguments are unpersuasive because the Horvitz teaching
of “caching data guided by a probabilistic predictive model” (Horvitz ¶ 9)—
i.e., ensuring that the content a user is most likely to view is available before
the user chooses to view the content—shows that there is a rational basis
underlying the Examiner’s proffered reason why it would have been obvious
to an artisan of ordinary skill to combine the teachings and suggestions of
Inoue and Horvitz. That is, rather than waste finite resources such as cache
Appeal 2019-004083
Application 14/911,079

12

space, network bandwidth, processing power, etc., storing content a user is
unlikely to view, Horvitz teaches or suggests using such resources for the
content the user will most likely view. See id.
Accordingly, we sustain the Examiner’s 35 U.S.C. § 103 rejection of
claim 8, and claims 9, 10, and 13–16, which Appellant does not argue
separately. See Appeal Br. 18.
Claim 113
The Examiner finds that Leighton—which teaches that a server can
elect to not serve an object when requests for the object exceed a certain
threshold (Leighton ¶ 74)—teaches or suggests, depending on a computer
workload, inhibiting or triggering of the claim 8 step of obtaining a first
recommendation list, automatically selecting at least one multimedia content
identifier, and pre-computing access data (Final Act. 17 (citing Leighton
¶¶ 60, 68, 74)).
Appellant contends the Examiner erred because “one of ordinary skill
in the art would have reasonably understood that” Leighton’s objects are
“embedded in a base HTML document” and the Examiner failed to provide
“any support for the proposition that one of ordinary skill in the art would
have understood Leighton as inhibiting service of content identifiers.”

3 Although tangential to our analysis, we note that claim 11 may not comply
with 35 U.S.C. § 112(d) as claim 8, from which claim 11 depends, performs
step “e),” but claim 11 recites “inhibiting . . . step e), depending on the
computed workload” (emphasis added). That is, claim 11 appears to
broaden, rather than further limit, the method of claim 8. In the event of
further prosecution, we recommend Appellant and the Examiner ensure that
claim 11 complies with 35 U.S.C. § 112(d).
Appeal 2019-004083
Application 14/911,079

13

Appeal Br. 19; see also Reply Br. 16 (the claimed “act of pre-computing
does not involve serving content”).
Appellant’s arguments are unpersuasive because the Examiner’s
findings show that it would have been obvious to an artisan of ordinary skill
to limit the services a server provides when a server’s workload exceeds a
particular threshold “to provide an infrastructure designed to serve content
efficiently, effectively, and reliably to end users.” Ans. 14 (citing Leighton
¶ 12). The Examiner properly relies on Inoue and Horvitz to teach or suggest
the claimed source of workload and services to limit when that threshold
exceeds a particular threshold. See id.; see also, e.g., Final Act. 13–14.
Accordingly, we sustain the Examiner’s 35 U.S.C. § 103 rejection of
claim 11.
Claims 17 and 18
The Examiner finds that Relan—which teaches enabling a user to
unsubscribe from programming channels (see, e.g., Relan ¶ 116)—teaches
or suggests, in response to receiving a list of identifiers of users
unsubscribed from the system, triggering the claim 8 step of obtaining a first
recommendation list, automatically selecting at least one multimedia content
identifier, and pre-computing access data (Final Act. 18 (citing Relan
¶¶ 116, 132, 219)). Specifically, the Examiner finds that Relan’s analysis of
a subscriber’s viewing habits, along with the subscriber’s ability to
unsubscribe from subscribed programming, teaches or suggests the use of
such information in tailoring recommendations to the user. See Ans. 12–13
(citing Relan ¶¶ 214–20).
Appellant contends the Examiner erred because Relan’s teaching of
“how a user who is tired of subscribing can unsubscribe . . . is completely
Appeal 2019-004083
Application 14/911,079

14

unrelated to . . . triggering any step, let alone a step with all the diverse
operations of step (e)” of independent claim 8. Appeal Br. 17; see also Reply
Br. 15–16.
Appellant’s arguments are unpersuasive, however, because the
Examiner’s findings persuasively show that a user unsubscribing from
content affects the likelihood that the user will view that content. See, e.g.,
Ans. 12. As the Examiner’s findings show, it would have been obvious to an
artisan of ordinary skill to enable updating the analysis of viewing habits and
interests—updating the recommendations determined and the access data
pre-computed—in response to an unsubscription because such as a response
would “provide . . . more accurate and updated recommendations.” Ans. 13.
Accordingly, we sustain the Examiner’s 35 U.S.C. § 103 rejection of
claim 17, and claim 18, which Appellant does not argue separately. See
Appeal Br. 17–18.
CONCLUSION
Claims
Rejected 35 U.S.C. § References Affirmed Reversed
8–10, 13–16 103 Inoue, Horvitz 8–10, 13–16
11 103 Inoue, Horvitz,
Leighton
11
17, 18 103 Inoue, Horvitz,
Relan
17, 18
Overall
Outcome
8–11, 13–18
Appeal 2019-004083
Application 14/911,079

15

TIME PERIOD FOR RESPONSE
No time period for taking subsequent action in connection with this
appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 1.136(a)(1)(iv).
AFFIRMED