Ex Parte Brobst

Patent Trial and Appeal Board

Feb 21, 2017

11281527 (P.T.A.B. Feb. 21, 2017)

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.
11/281,527 11/17/2005 Stephen Brobst 11044.C02 2544
26890 7590 02/23/2017
TA1UFS M STOVFR
EXAMINER
TERADATA US, INC. CAO, PHUONG THAO
10000 INNOVATION DRIVE
DAYTON, OH 45342 ART UNIT PAPER NUMBER
2164
NOTIFICATION DATE DELIVERY MODE
02/23/2017 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):
michelle.boldman @ teradata. com
j ames. stover @ teradata.com
td.uspto@outlook.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte STEPHEN BROBST
Appeal 2016-006544
Application 11/281,527
Technology Center 2100
Before JOSEPH L. DIXON, ST. JOHN COURTENAY III, and
CARLA M. KRIVAK, Administrative Patent Judges.
DIXON, Administrative Patent Judge.
DECISION ON APPEAL
Appeal 2016-006544
Application 11/281,527
STATEMENT OF THE CASE
Appellant appeals under 35 U.S.C. § 134(a) from a rejection of claims
1—5, 8—12, 15—20, and 23—27. Claims 6, 7, 13, 14, 21, and 22 are cancelled.
This appeal is related to prior Appeal 2010-012525 of this application decided
May 20, 2013, where we reversed the Examiner’s obviousness rejection. We
have jurisdiction under 35 U.S.C. § 6(b).
We reverse.
The claims are directed to a self-adjusting database-query optimizer.
Claim 1, reproduced below, is illustrative of the claimed subject matter:
1. A computer-implemented method for use in executing a
query in a database system, the method comprising:
formulating an initial query-execution plan that identifies
an expected path for execution of the query;
formulating an alternative query-execution plan for
execution of the query;
initiating execution of the query according to said initial
query-execution plan; and
at some point after execution of the query has begun:
concluding that execution of the query has not proceeded
along the expected path; and
in response, choosing said alternative query-execution
plan for execution of the query.
2
Appeal 2016-006544
Application 11/281,527
REFERENCES
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Bestgen et al. US 2003/0018618 A1 Jan. 23,2003
V. Markl et al., “LEO: An autonomic query optimizer for DB2,” IBM
SYSTEMS JOURNAL, Vol. 42, No. 1, pp 98-106 (2003).
Kabra et al., “Efficient Mid-Query Re-optimization of Sub-optimal
Query Execution Plans,” ACM SIGMOD Record, pp. 106—117
(1998).
REJECTIONS
The Examiner made the following rejections:
Claims 1, 2, 9, 12, 16, 17, and 24 stand rejected under 35 U.S.C.
§ 102(b) as being anticipated by Markl. Final Act. 3—11.
Claims 3—5, 18—20, and 27 stand rejected under 35 U.S.C. § 103(a) as
being unpatentable over Markl in view of Bestgen. Final Act. 11—17.
Claims 8, 10, 11, 15, 23, 25, and 26 stand rejected under 35 U.S.C.
§ 103(a) as being unpatentable over Markl in view of Kabra. Final Act. 18—
28.
ANALYSIS
Appellants address independent claims 1, 12, 16, and 27 together as a
group. (App. Br. 8). We select independent claim 1 as the illustrative claim
for the group. We recognize that independent claim 27 is rejected under
obviousness rather than anticipation due to its additional detailed limitations.
3
Appeal 2016-006544
Application 11/281,527
Anticipation
“It is well settled that a prior art reference may anticipate when the
claim limitations not expressly found in that reference are nonetheless
inherent in it. Under the principles of inherency, if the prior art necessarily
functions in accordance with, or includes, the claimed limitations, it
anticipates.” In re Cruciferous Sprout Litig., 301 F.3d 1343, 1349 (Fed. Cir.
2002) (citations and internal quotation marks omitted). “Inherency,
however, may not be established by probabilities or possibilities. The mere
fact that a certain thing may result from a given set of circumstances is not
sufficient.” In re Robertson, 169 F.3d 743, 745 (Fed. Cir. 1999) (citations
and internal quotation marks omitted).
With respect to independent claim 1, Appellants generally contend the
Markl reference does not teach or suggest the specific invention as recited in
the steps of independent claim 1. (App. Br. 8—9). Appellants specifically
contend:
[MJonitoring and comparing optimizer estimates with actual
cardinalities at each step in a query execution plan, and computes
adjustments to its estimates that may be used during future
optimizations of similar queries. Markl at page 98, left column,
lines 22-30 further states that the detection of estimation errors
can also trigger reoptimization of a query in mid-execution. It
is not seen that either of these functions - computing adjustments
for use during future optimizations of similar queries, or
reoptimization of a query in mid-execution - is equivalent to,
teaches, or suggests the steps of ‘initiating execution of the query
according to said initial query-execution plan’ and ‘at some point
after execution of the query has begun: teach that execution of
the query has not proceeded along the expected path; and in
response, choosing said alternative query-execution plan for
execution of the query’ recited in independent claim 1, and
similar limitations recited in independent claims 12, 49, and 27.
4
Appeal 2016-006544
Application 11/281,527
Markl does not teach having an alternative query execution plan
which is executed in place of the initial query execution when
the initial execution plan has not proceeded along the expected
path, as recited in independent claims 1, 12, 16, and 27.
(App. Br. 9-10).
The Examiner responds to Appellants’ contentions regarding the
Examiner’s reliance upon pages 98 and 100 of the Markl reference as set
forth in the statement of the rejection in the Examiner’s Answer. The
Examiner further provides additional proffered support for the Markl
reference anticipating independent claim 1. The Examiner identifies
additional disclosures on pages 100 and 103 of the Markl reference. (Ans.
3—4).
Giving the claim a broad yet reasonable interpretation, we agree with
the Examiner the Markl reference discloses the invention as set forth in
independent claim 1 with the exception of the last limitation.
“[Although a method claim necessarily recites the steps of the
method in a particular order, as a general rule the claim is not limited to
performance of the steps in the order recited, unless the claim explicitly or
implicitly requires a specific order.” Baldwin Graphic Sys., Inc. v. Siebert,
Inc., 512 F.3d 1338, 1345 (Fed. Cir. 2008).
Appellant’s arguments imply that claim 1 requires the step of
“formulating an alternative query-execution plan for execution of the query”
must be performed before the “initiating execution of the query” step. (App.
Br. 8—10). We disagree with Appellant and conclude that under a broad but
reasonable interpretation, the step of “formulating an alternative query-
execution plan” may be performed at any time prior to the last recited step
5
Appeal 2016-006544
Application 11/281,527
of “in response, choosing said alternative query-execution plan for execution
of the query.” (Claim 1). See Baldwin Graphic Sys., Inc., 512 F.3d at 1345.
Consequently, Appellant’s argument regarding the formulating step
does not distinguish from the decision to re-optimize and the choice of the
re-optimization of the query as disclosed on pages 98 and 101 of the Markel
reference.
Appellant additionally contends:
Markl: page 99, left column, second paragraph, lines 1-5,
was cited as teaching the step “formulating an alternative query-
execution plan for execution of the query” recited in independent
claim 1, and similar limitations recited in independent claims 12,
16, and 27.
Markl at page 99, left column, second paragraph, lines 1 -
5 provides ‘Most modem query optimizers determine the best
query execution plan (QEP, or simply plan) for executing an
SQL query by mathematically modeling the execution cost for
each of many alternative QEPs and choosing the one with the
lowest estimated cost.’ Markl at page 99, left column, second
paragraph, lines 1-5 describes choosing a single query execution
plan from many alternative plans. However, Markl does not
teach selecting an alternative plan to be executed in place of the
initial query execution when the initial execution plan has not
proceeded along the expected path, as recited in independent
claims 1, 12, 16, and 27.
(App. Br. 9). We agree with Appellant that Markl does not teach having an
alternative query execution plan which is executed in place of the initial
query execution when the initial execution plan has not proceeded along the
expected path, as recited in independent claims 1, 12, 16, and 27. (App. Br.
8—10; Reply Br. 3).
Appellant further contends:
6
Appeal 2016-006544
Application 11/281,527
Markl at page 101, right column, third full paragraph, in
describing immediate feedback-based lear[n]ing, states ‘As part
of the LEO project, we are currently investigating how to use
this knowledge immediately by dynamically reoptimizing the
current query and changing its execution plan, if all of the rows
for an intermediate result are materialized before proceeding at
any point in the plan. ’ Thus, at time of publication, Markl fails
to present a solution which dynamically reoptimizes a current
query and changes its execution plan, if all of the rows for an
intermediate result are materialized before proceeding at any
point in the plan, i.e., Markl does not teach having an alternative
query execution plan which is executed in place of the initial
query execution when the initial execution plan has not
proceeded along the expected path, as recited in independent
claims 1, 12, 16, and 27.
Markl, at page 103, right column, first full paragraph
beginning, ‘When to optimize,’ states ‘immediate learning can
change the plan for a query at run time, when the actual
cardinalities are significantly different from the estimated
cardinalities.’ Markl, at page 103, describes an approach for
changing a query plan at run time, not when the initial execution
plan has not proceeded along the expected path, as recited in
independent claims 1, 12, 16, and 27.
Applicant would like to note that optimizing, or
reoptimizing, the plan, differs from executing the plan. Much of
the language cited by the Exa[]m[i]ner in the rejection of the
claims concerns query plan optimization or reoptimization and
when the optimizations are performed. It is not seen that Markl
teaches having an alternative query execution plan which is
executed in place of the initial query execution when the initial
execution plan has not proceeded along the expected path, as
recited in independent claims 1, 12, 16, and 27.
(Reply Br. 3—4).
Appellant timely responds to the Examiner’s further clarifications and
additional reliance as set forth in the Response to Arguments section of the
Examiner’s Answer in the Reply Brief. (Reply Br. 2—A). Appellant
7
Appeal 2016-006544
Application 11/281,527
repeatedly contends the Markl reference does not disclose having an
alternative query execution plan, which is executed in place of the initial
query execution plan, when the initial execution plan is not proceeding along
the expected path. (Reply Br. 3). We agree with Appellant.
We note that the Board is a reviewing body and not a place of initial
examination. Moreover, it is our view that the rigorous requirements of 35
U.S.C. § 102 essentially require a one-for-one mapping of each argued
limitation to the corresponding portion of the reference, which the Examiner
must identify with particularity. Here, we find the Examiner’s anticipation
rejection is not supported by the express or inherent disclosure of Markl.
To affirm the Examiner on this record would require us to resort to
speculation or unfounded assumptions. See In re Warner, 379 F.2d 1011,
1017 (CCPA 1967). We decline to engage in speculation. Consequently,
we cannot sustain the rejection of independent claim 1, and independent
claims 12 and 16, which each recite the contested limitations in
commensurate form
Obviousness
With respect to independent claim 27, the Examiner has not identified
additional teachings in the Bestgen reference to remedy the noted deficiency
above with respect to the execution of the alternative query execution plan in
place of the initial query execution plan. As a result, we cannot sustain the
obviousness rejection of independent claim 27 for the same reasons
addressed above.
With respect to dependent claims 3—5, 8, 10, 11, 15, 18—20, 23, 25,
and 26, Appellants contend the Bestgen and Kabra references were cited to
teach specific additional limitations recited in the dependent claims and the
8
Appeal 2016-006544
Application 11/281,527
Examiner has not shown that these references remedy the defects noted
above with respect to the Markl reference. Consequently, we cannot sustain
the obviousness rejections of claims 3—5, 8, 10, 11, 15, 18—20, 23, 25, and
26 for the same reasons addressed above.
CONCLUSIONS
The Examiner erred in rejecting claims 1,2, 9, 12, 16, 17, and 24
based upon anticipation under 35 U.S.C. § 102. The Examiner also erred in
rejecting claims 3—5, 8, 10, 11, 15, 18—20, 23, and 25—27 based upon
obviousness under 35 U.S.C. § 103.
DECISION
For the above reasons, we reverse the Examiner’s anticipation and
obviousness rejections of claims 1—5, 8—12, 15—20, and 23—27.
REVERSED
9