SonicWALL, Inc.

Patent Trials and Appeals Board

Mar 6, 2020

13853946 - (D) (P.T.A.B. Mar. 6, 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.
13/853,946 03/29/2013 Thomas E. Raffill SWUS-058REI 4033
138555 7590 03/06/2020
Polsinelli LLP - SONICWALL US HOLDINGS INC.
3 Embarcadero Center
Suite 2400
San Francisco, CA 94111
EXAMINER
WASSUM, LUKE S
ART UNIT PAPER NUMBER
3992
NOTIFICATION DATE DELIVERY MODE
03/06/2020 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):
sfpatent@polsinelli.com
uspt@polsinelli.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________

Ex parte THOMAS E. RAFFILL, SHUNHUI ZHU, ROMAN
YANOVSKY, BORIS YANOVSKY, and JOHN GMEUNDER
____________

Appeal 2020-001587
Reissue Application 13/853,946
Patent 7,917,522 B1
Technology Center 3900
____________

Before JOHN A. JEFFERY, ERIC B. CHEN, and JENNIFER L.
McKEOWN, Administrative Patent Judges.

JEFFERY, Administrative Patent Judge.

DECISION ON APPEAL
Under 35 U.S.C. § 134(a), Appellant1 appeals from the Examiner’s
decision to reject claims 1–10. We have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

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 SonicWALL
US Holdings Inc. Appeal Br. 3.
Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

2

STATEMENT OF THE CASE
Appellant seeks to reissue U.S. Patent 7,917,522 B2 (“’522 patent”)
directed to building a statistical model for content classification. In one
aspect, a threshold for each of plural categories based on (1) validation
document pre-classification; (2) occurrence of N-grams in validation
documents; (3) N-gram-based scores; and (4) a predetermined false positive
limit. See generally ’522 patent, col. 1, ll. 14–16; col. 4, ll. 26–67; col. 5, l.
28 – col. 6, l. 10; Figs. 1A–3. Claim 1 is illustrative:

1. A computer-implemented method, comprising:
generating a plurality of scores for each of a plurality of
N-grams with respect to each of a plurality of categories based
on occurrence of the plurality of N-grams in a plurality of
training documents;
setting a threshold for each of the plurality of categories
based on preclassification of a plurality of validation
documents, occurrence of the plurality of Ngrams in the
plurality of validation documents, the plurality of scores, and a
predetermined false positive limit;
generating a statistical content classification model based
on the plurality of scores and the threshold;
assigning a plurality of weights for each of the plurality of N-
grams, wherein each of the plurality of weights is associated
with a distinct one of the plurality of categories;
storing the statistical content classification model in a
data storage device; and
allowing content filters to access the statistical content
classification model in the data storage device over a network,
wherein the content filters classify contents into one of the
plurality of categories using the statistical content classification
model.

Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

3
THE REJECTIONS
The Examiner rejected claims 1–10 under 35 U.S.C. § 251 as based
on a defective reissue declaration. Final Act. 8–9.2
The Examiner rejected claims 1–10 under 35 U.S.C. § 103 as
unpatentable over Guilak (US 2003/0225763 Al; published Dec. 4, 2003),
Alspector (US 7,725,475 Bl; issued May 25, 2010), and Gardner (US
2006/0053174 Al; published Mar. 9, 2006). Final Act. 10–45.

THE § 251 REJECTION
Because Appellant does not contest the Examiner’s § 251 rejection of
claims 1–10 (Final Act. 8–9; Ans. 4), we summarily sustain this rejection.
See Manual of Patent Examining Procedure (MPEP) § 1205.02 (9th ed. Rev.
08.2017, Jan. 2018) (“If a ground of rejection stated by the examiner is not
addressed in the appellant’s brief, appellant has waived any challenge to that
ground of rejection and the Board may summarily sustain it, unless the
examiner subsequently withdrew the rejection in the examiner’s answer.”).

THE OBVIOUSNESS REJECTION
The Examiner finds that Guilak discloses every recited element of
independent claim 1 including setting a threshold for each category, but does
not do so based on (1) validation document pre-classification; (2) occurrence
of N-grams in the validation documents; (3) the recited scores; and (4) a

2 Throughout this opinion, we refer to (1) the Final Rejection mailed March
28, 2019 (“Final Act.”); (2) the Appeal Brief filed August 2, 2019 (“Appeal
Br.”); (3) the Examiner’s Answer mailed October 23, 2019 (“Ans.”); and (4)
the Reply Brief filed December 23, 2019 (“Reply Br.”).
Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

4
predetermined false positive limit. Final Act. 10–14. The Examiner,
however, cites Alspector for teaching the first three bases, and Gardner for
teaching basis (4) in concluding that the claim would have been obvious
over the cited references’ collective teachings. Final Act. 12–15.
Appellant argues that the Examiner’s reliance on Gardner for teaching
setting a threshold for each category based on a predetermined false positive
limit is misplaced because, among other things, Gardner’s rules engine and
associated rules do not pertain to a document classifier. Appeal Br. 9–13;
Reply Br. 6–14. According to Appellant, Gardner’s rule constraints are not
equivalent to the recited threshold, nor does Gardner set or generate a
threshold, let alone do so based on a false positive limit as claimed. Appeal
Br. 11–13; Reply Br. 6–14.

ISSUE
Under § 103, has the Examiner erred in rejecting claim 1 by finding
that Guilak, Alspector, and Gardner collectively would have taught or
suggested setting a threshold for each category based on, among other
things, a predetermined false positive limit (“the false positive limitation”)?

ANALYSIS
We begin by noting that the Examiner’s reliance on the teachings of
Guilak and Alspector in connection with the particular limitations on which
they were relied is undisputed, including the Examiner’s reliance on Guilak
for teaching setting a threshold for each category, albeit lacking the four
particular recited techniques for doing so. See Final Act. 10–14; Ans. 18,
Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

5
20. Nor does Appellant dispute the cited references’ combinability. Rather,
as noted above, this dispute turns solely on the Examiner’s reliance on
Gardner for teaching the fourth recited basis on which a threshold is set for
each category, namely that the threshold is set based on a predetermined
false positive limit. Therefore, we confine our discussion principally to
Gardner.
On this record, we see no error in the Examiner’s reliance on Gardner
for at least suggesting the false positive limitation. See Final Act. 14–15;
Ans. 14–18. As noted in the Abstract, Gardner’s system extracts and
manages data to create multi-relational ontologies, where extraction can
involve rules and natural-language processing. To this end, Gardner’s
system can use a “rules engine” that enables creating, organizing, validating,
modifying, storing, and/or applying various rules to create, maintain, and use
ontologies. Gardner ¶ 120. These rules can pertain to, among other things,
linguistic analysis, assertion extraction, curation, semantic normalization,
and inference. Id. As Gardner explains, applying rules to a corpus of
documents may generate rule-based products including, for example, tagged
document content, rule-based assertions, or “other product information.” Id.
A key aspect of Gardner’s rules engine is that it can validate rules by
comparing the actual results of applying a rule to expected results. See
Gardner ¶ 123. This validation produces information regarding an
individual rule’s quality in terms of, for example, the percentage of false
positives returned by a particular rule. Id.
To validate rules, Gardner’s rules engine may use predetermined
thresholds for false positives and negatives. Id. ¶ 124. For example, if the
Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

6
percentage of false positives exceeds predetermined thresholds for a
particular rule, that rule may be deleted, replaced, or modified, the latter
including “tightening” the rule’s constraints to reduce or eliminate
unexpected results. Id.
The clear import of this functionality is that by “tightening” a rule’s
constraints responsive to the percentage of false positives exceeding a
predetermined threshold, such that this “tightening” reduces or eliminates
unexpected results, Gardner at least suggests setting the threshold
accordingly. Although Gardner may not state explicitly that this threshold
adjustment is for each of plural categories as claimed, we nonetheless see no
reason why such a threshold adjustment technique could not be provided in
connection with Guilak’s system that sets thresholds for each of plural
categories as shown in in step 612 of Figure 6 as the Examiner proposes to,
among other things, increase category-based classification accuracy—a
predictable result. See Final Act. 14–15; Ans. 6–11 (citing Guilak ¶ 55).
Such an enhancement uses prior art elements predictably according to their
established functions—an obvious improvement. See KSR Int’l Co. v.
Teleflex, Inc., 550 U.S. 398, 417 (2007). To the extent that Appellant
contends that applying a threshold adjustment technique, such as that
disclosed in Gardner, to Guilak’s category-based threshold setting technique
as the Examiner proposes would have been uniquely challenging or
otherwise beyond the level of ordinarily skilled artisans, there is no
persuasive evidence on this record to substantiate such a contention.
Appellant’s arguments regarding Gardner’s individual shortcomings in this
regard (see Appeal Br. 10–13; Reply Br. 7–12) do not show nonobviousness
Appeal 2020-001587
Application 13/853,946
Patent 7,917,522 B1

7
where, as here, the rejection is based on the cited references’ collective
teachings. See In re Merck & Co., 800 F.2d 1091, 1097 (Fed. Cir. 1986).
Therefore, we are not persuaded that the Examiner erred in rejecting
claim 1, and claims 2–10 not argued separately with particularity.

CONCLUSION
In summary:
Claims
Rejected
35 U.S.C.
§
Reference(s)
/Basis
Affirmed Reversed
1–10 251 Defective
Reissue
Declaration
1–10
1–10 103 Guilak,
Alspector,
Gardner
1–10
Overall
Outcome
1–10

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).

AFFIRMED