Ex Parte Yaros et al

Patent Trial and Appeal Board

Jul 26, 2018

14218544 (P.T.A.B. Jul. 26, 2018)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO.
14/218,544
152606 7590
VMWare-OPW
P.O. Box 4277
Seattle, WA 98194
FILING DATE FIRST NAMED INVENTOR
03/18/2014 AlYaros
07/26/2018
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
B556 1555
EXAMINER
DALENCOURT, YVES
ART UNIT PAPER NUMBER
2457
MAIL DATE DELIVERY MODE
07/26/2018 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 ALY AROS, EY AL COHEN, EVGENY ETKIN,
and ASAP ABRAMOVITZ
Appeal2018-001932
Application 14/218,544 1
Technology Center 2400
Before CARLA M. KRIVAK, HUNG H. BUI, and JON M. JURGOV AN,
Administrative Patent Judges.
JURGOV AN, Administrative Patent Judge.
DECISION ON APPEAL
Appellants seek review under 35 U.S.C. § 134(a) from a Final
Rejection of claims 1-3, 6-8, and 11-13. Claims 4, 5, 9, 10, 14, and 15 are
indicated as containing allowable subject matter (Final Act. 6). We have
jurisdiction under 35 U.S.C. § 6(b ).
We affirm. 2
1 Appellants identify VMWARE, INC. as the real party in interest.
(App. Br. 1.)
2 Our Decision refers to the Specification ("Spec.") filed March 18, 2014,
the Final Office Action ("Final Act.") mailed July 20, 2016, the Appeal
Brief ("App. Br.") filed January 18, 2017, the Examiner's Answer ("Ans.")
mailed May 3, 201 7, and the Reply Brief ("Reply Br.") filed September 7,
2017.
Appeal2018-001932
Application 14/218,544
CLAIMED INVENTION
The claims are directed to methods and systems "for detecting and
correcting, or deleting, data anomalies in data generated by information
technology business management ('ITBM') systems." (Abstract.)
Appellants' invention "detects one or more data anomalies in [a] record of
data based on the numerical values of the [record's] data types in the
[record's] time periods" and "enables a user to correct the data anomalies in
the recorded data based on the user's decision to selectively correct or delete
each of the data anomalies." (Abstract.)
Claims 1, 6, and 11 are independent. Claim 1, reproduced below, is
illustrative of the claimed subject matter:
1. A system for correcting data anomalies comprising:
one or more processors;
one or more data-storage devices; and
machine-readable instructions stored in the one or more
data-storage devices that when executed using the one or more
processors controls the system to perform
receiving a record of data output from an adaptor
or a data-management operator of an information
technology business management system, the record of
data including data types and associated numerical values
recorded over a number of time periods;
detecting one or more data anomalies in the record
of data based on the numerical values of the data types in
the time periods;
reporting a set of the one or more data anomalies;
and
correcting the data anomalies in the recorded data
by deleting each of the data anomalies or by replacing
each of the data anomalies with one of a mean value,
2
Appeal2018-001932
Application 14/218,544
median value, minimum value, or a maximum value of
other numerical values of the same data type.
(App. Br. 14 (Claims App'x).)
REJECTION & REFERENCES
Claims 1-3, 6-8, and 11-13 stand rejected under 35 U.S.C. § 103
based on Dragoljub Pokrajac et al., Incremental Local Outlier Detection for
Data Streams, PROCEEDINGS OF THE 2007 IEEE SYMPOSIUM ON
COMPUTATIONAL INTELLIGENCE AND DATA MINING 504--15 (2007)
("Pokrajac") and Maeda et al. (US 2014/0195184 Al, published July 10,
2014) ("Maeda"). (Final Act. 3-7.)
ANALYSIS
Appellants contend Pokrajac does not teach or suggest "data types and
time periods," and "receiving a record of data ... including data types and
associated numerical values recorded over a number of time periods," as
recited in claim 1. (Reply Br. 2-3; see also App. Br. 6.) Particularly,
Appellants argue Pokrajac does not disclose "data types and time periods
associated with data streams the 'Periodic' and 'Supervised' LOP [(Local
Outlier Factor)] algorithms are applied to," and "does not teach or suggest
that the data block, [to which] the 'Periodic' LOP algorithm is applied to,
includes 'data types' and 'numerical values recorded of a number of time
periods."' (Reply Br. 2-3 ( citing Pokrajac p. 505 (§ II. Background), Figs.
1-2).)
We do not find Appellants' arguments persuasive. Instead, we find
the Examiner has provided a comprehensive response to Appellants'
3
Appeal2018-001932
Application 14/218,544
arguments supported by a preponderance of evidence. ( Ans. 2-7.) As such,
we adopt the Examiner's findings and explanations provided therein. (Id.)
At the outset, we note claim terms are given their broadest reasonable
interpretation consistent with the Specification. In re Am. Acad. of Sci. Tech
Ctr., 367 F.3d 1359, 1364 (Fed. Cir. 2004). Under the broadest reasonable
interpretation, claim terms are given their ordinary and customary meaning,
as would be understood by one of ordinary skill in the art in the context of
the entire disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257
(Fed. Cir. 2007). At the same time, care must be exercised not to import
limitations into the claims or to read a particular embodiment appearing in
the written description into the claim if the claim language is broader than
the embodiment. In re Van Geuns, 988 F.2d 1181, 1184 (Fed. Cir. 1993)
(citing In re Zietz, 893 F.2d 319,321 (Fed. Cir. 1989)).
Appellants' claim 1 recites, inter alia, "a record of data" that includes
"data types" and "associated numerical values recorded over a number of
time periods." (App. Br. 14 (Claims App'x).)
Appellants' Specification does not provide explicit and exclusive
definitions of the claim terms "record of data," "data types," "numerical
values," and "time periods." Rather, Appellants' Specification provides
discussion of non-limiting examples of the claim terms. For example, the
Specification describes "a record of data" with "data extracted from one or
more data sources," the record of data ( 400) including:
a data type column 402, a numerical value column 403, a
description column 404, and a column identified as other 405.
Data table entries for data types, numerical quantities,
descriptions, and other depends on the kind of data collected,
manipulated, and organized to form the record 400. Each row,
such as row 406, identifies a specific data type 408, a numerical
4
Appeal2018-001932
Application 14/218,544
value 409 associated with the data type 408, a description 410
of the data type 408, and other 411 information regarding the
data type identified in entry 408.
(Spec. ,r 23 ( emphases added); see also Spec. ,r,r 18, 24, 25, Fig. 4.) Thus,
Appellants' Specification broadly describes "data types" are kinds or
categories of data in a record of data. (See Spec. ,r,r 23-25, 34, Figs. 4---6
(emphasis added).) Data types/categories depend "on the kind of data
collected, manipulated, and organized to form the record [ of data]," and may
include "email accounts" and categories of company "expenses" such as
facilities and hardware expenses. (See Spec. ,r,r 23-25, 33-34, Figs. 4---6.)
The Specification further describes "numerical values" are quantities or
values associated with a "data type," such as an "amount of storage in GBs"
("numerical values") for particular "email accounts" ("data types"), and
allocation "percentages" and "costs" ("numerical values") expended for
facilities or hardware ("data types"). (See Spec. ,r,r 23-25, 28, 33-34
( emphasis added).) The Specification also describes "time periods"
designating times at which "numerical values" have been collected or
recorded, the time periods including "an interval of time over which data
storage for each email account is recorded," "a billing cycle or interval of
time between billing the enterprise for email storage," and a "period of time
over which the services listed in [a] bill of IT ... were used by the
enterprise." (See Spec. ,r,r 24--25, 27-29, 34 (emphasis added).) The "time
periods" may include "a current batch of time periods," a "current month,"
"a much larger interval of time" including "any number of time periods
preceding the recent time periods," and "any number of previous months."
(See Spec. ,r 27 .)
5
Appeal2018-001932
Application 14/218,544
We, therefore, agree with the Examiner that Pokrajac's "real life data
sets," including image data and video streams' data from surveillance
applications, are commensurate with the broad description of record of data
in Appellants' Specification. (Final Act. 3--4 (citing Pokrajac pp. 504--505);
Ans. 2-3 (citing Pokrajac pp. 505-506)3; see Spec. ,r,r 18, 23-25.)
Pokrajac's record of data includes "real life data," such as image data, video
frames, and video trajectories extracted from video streams for processing
by LOP algorithms. (Final Act. 3--4 ( emphasis omitted); see Pokrajac p. 504
(see also pp. 512-513) (describing "real life data sets" with "video frames"
and "video motion trajectories").) Pokrajac's video frames and video
motion trajectories, which indicate data categories, are commensurate with
the broad description of data types in Appellants' Specification. (Final Act.
3--4; see Spec. ,r,r 23-25, 34.) Further, Pokrajac discloses "data records
(samples)" of the video frames and video motion trajectories are described
numerically (for processing by the LOP algorithms), thereby teaching that
numerical values are associated with the data types, as claimed. (Final Act.
3--4; Ans. 2-3; see Pokrajac p. 505 (see also pp. 512-513) (disclosing video
motion trajectories "represented by ... points in [x,y,time] space (two spatial
coordinates on the frame and the time instant)").) Additionally, Pokrajac
discloses the data records/samples of the video frames and trajectories
("numerical values" associated with the "data types") are recorded in
"block[ s] of 1000 data records" or in sequential "time interval[s]," which is
commensurate with the broad description of numerical values recorded over
3 Portions of Pokrajac quoted on page 2 in the Examiner's Answer are from
Pokrajac's pages 505 and 506.
6
Appeal2018-001932
Application 14/218,544
a number of time periods in Appellants' Specification. (Ans. 2-3 (citing
Pokrajac pp. 505-506); Final Act. 3--4; see Spec. ,r,r 24--25, 27-29, 34.)
Thus, we agree with the Examiner that Pokrajac teaches receiving a
record of data including data types and associated numerical values recorded
over a number of time periods, as recited in claim 1.
Next, Appellants argue Pokrajac does not teach or suggest "detecting
one or more data anomalies in the record of data based on the numerical
values of the data types in the time periods," as recited in claim 1. (Reply
Br. 4; App. Br. 6-7.) Appellants argue "[t]here is no mention [in Pokrajac]
of data types or a number of time periods," and the Examiner has not
provided an explicit analysis that explains where in Pokrajac "one or more
data anomalies [ are detected] in the record of data based on the numerical
values of the data types in the time periods." (Reply Br. 4; App. Br. 6-7.)
We do not agree.
Our review of Appellants' Specification indicates that detecting
anomalies in a record of data based on numerical values of data types in the
pertinent time periods is performed by detecting outliers-"anomalous
numerical data associated with one or more data types"----of a set of k-
nearest neighbors in the record of data. (See Spec. ,r,r 27-28.) Examples of
anomaly detection in the Specification include "detect[ing] data storage
anomalies for any single email account over a number of recent time
periods" and detecting "that in one of the billing cycles the cost associated
with the expense 'facilities' ... [is] much higher or lower than the costs
associated with the expense 'facilities' for all other billing cycles recorded in
the data." (See Spec. ,r,r 27, 33.) Pokrajac's outlier detection algorithms
similarly detect anomalies by "determin[ing] outliers once all the data
7
Appeal2018-001932
Application 14/218,544
records (samples) are present in the dataset" or by "identify[ing] outliers as
soon as new data record appears in the dataset." (Ans. 4 (citing Pokrajac p.
505)4; see also Final Act. 4 (citing Pokrajac pp. 504--505 (providing that
"outliers (rare events)" are "distinct from all other data records" in the data
stream).) Thus, Pokrajac's outlier detection is commensurate with the
description in the Specification of detecting anomalies based on numerical
values of data types in relevant time periods. (Final Act. 4; Ans. 3--4.)
Therefore, we agree with the Examiner that Pokrajac teaches
"detecting one or more data anomalies in the record of data based on the
numerical values of the data types in the time periods" as recited in claim 1.
Appellants also argue Pokrajac does not teach or suggest "reporting a
set of the one or more data anomalies," or the claimed order of receiving,
detecting, and reporting steps. (Reply Br. 4--5, 8; App. Br. 7-8.) We do not
agree, because Pokrajac explains it is important for an analyst to understand
the difference between normal and unusual behavior in streams of data, and
develops outlier detection algorithms that enable "incremental
outlier/anomaly detection that can adapt to novel behavior and provide
timely identification of unusual events." (Ans. 4--5 (citing Pokrajac p.
504)5; Final Act. 4; see also Pokrajac p. 513 (explaining that an incremental
LOP algorithm finds unusual trajectories in surveillance videos which "may
lead to early identification of illegal behavior").) Thus, Pokrajac teaches
reporting anomalies that have been detected in a record of data, as required
by claim 1. (Final Act. 4; Ans. 5.)
4 Portions of Pokrajac quoted on page 4 in the Answer are from Pokrajac's
page 505. (See Ans. 4.)
5 Portions of Pokrajac quoted on pages 4 and 5 in the Answer are from
Pokrajac's page 504. (See Ans. 4--5.)
8
Appeal2018-001932
Application 14/218,544
Next, Appellants acknowledge the secondary reference to Maeda
discloses "various types of data [(e.g., sensor data)] used to detect anomalies
in the behavior of plant equipment," but argue "Maeda does not teach,
suggest, or mention correcting data anomalies in the signals output from the
sensors by deleting or replacing each of the data anomalies" as required by
claim 1. (App. Br. 8, 10 (emphasis added); see also Reply Br. 5.)
Appellants also argue "[t]here is no teaching or suggestion [in Maeda] of
replacing data anomalies with any of' a "mean value, median value,
minimum value, or a maximum value of other numerical values of the same
data type," as claimed. (App. Br. 10 (emphasis added).)
At the outset, we note Appellants' claim 1 is broadly worded to
include alternative claim limitations. For example, Appellants' claim 1
recites:
correcting the data anomalies in the recorded data by deleting
each of the data anomalies or by replacing each of the data
anomalies with one of a mean value, median value, minimum
value, or a maximum value of other numerical values of the
same data type.
(App. Br. 14 (Claims App'x) (emphases added).) Appellants' Specification
also describes the "deleting" and "replacing" actions in the alternative. 6 As
such, we agree with the Examiner that claim 1 's correcting step is performed
6 For example, Appellants' Specification provides "[t]he method also
enables a user to correct the data anomalies based on a user's decision to
selectively correct or delete each of the data anomalies." (Spec. ,r 4
( emphasis added).) For example, "[ w ]hen data types with anomalous values
are detected, a user may decide to replace the anomalous numerical values
with the mean or median of the numerical values over a number of time
periods for the same data type or the user may decide to delete the all data
associated with the anomalous value." (Spec. ,r 32 ( emphasis added).)
9
Appeal2018-001932
Application 14/218,544
[A] by deleting or [BJ by replacing, and claim 1 does not require "correcting
the data anomalies" to be performed by both "deleting each of the data
anomalies" and "replacing each of the data anomalies." (Ans. 6.)
Additionally, we agree with the Examiner that Maeda teaches at least one of
these correcting actions-particularly, the "deleting each of the data
anomalies." (Ans. 5---6.)
In particular, Maeda discloses detecting anomalies in a record of data,
which includes categories of sensor data (maintenance history information
from plant equipment) such as alarms, picture qualities, setting dial values,
resistance values, and set time values. (See Maeda ,r,r 83, 110 (disclosing
numerical values for sensor data); Final Act. 4 (citing Maeda ,r 110); Ans. 5
(citing Maeda ,r 83).) Sensor data anomalies are detected based on normal
operating curves and past examples of normal operation, and the detected
anomalies are used to "update and correct" a set of learning data in real
time. (See Ans. 5---6 (citing Maeda ,r,r 145, 147-148, 161)7; see also Final
Act. 4 (citing Maeda ,r 134); Maeda ,r,r 28, 83, 113, 134, 145, 147-148, 161.)
Specifically, Maeda "update[ s] and correct[ s] the learning data of
an anomaly sign" for specific plant equipment, to obtain "learning data
composed of mainly normal examples." (See Maeda ,r,r 147, 161; see also
Maeda ,r,r 145, 148 ("the learning data itself is subjected to a selection
process of taking or discarding the data. In this way, the learning data is ...
updated in order to improve the precision"); Ans. 6.) Maeda corrects
learning data for specific equipment by analyzing the equipment's sensor
7 We note pages 5 and 6 in the Examiner's Answer (stating that "Prokrajac
[sic] discloses") actually refer to paragraphs of Maeda, not of Pokrajac
(which does not include such paragraph numbers). (See Ans. 5---6.)
10
Appeal2018-001932
Application 14/218,544
data, and removing anomalous data from sensor data to obtain "learning data
[that] is data stored in the learning-data storage section 915 as data
composed of mainly normal examples." (See Maeda ,r 148; see also Maeda
,r,r 115-116 ("FIG. 4C is a block diagram for a learning time" illustrating
operations to create a recognition rule 443 and feature extraction
classifications 442 "in accordance with a phenomenon enlightening an
anomaly sign at a learning time by carrying out a segment cutting out
process[] 441 ... inputting sensor data 310 and making use of event data
105 and in accordance with countermeasure information 444 (part
replacement, adjustment, resumption and others)"), Fig. 4C.)
Therefore, we agree with the Examiner that Maeda teaches correcting
anomalies in recorded data by deleting each of the data anomalies, as
required by claim 1. (Ans. 5---6; Final Act. 4--5.)
Appellants further argue "[t]he Examiner ... failed to demonstrate a
reasonable expectation of success in combining Pokrajac with Maeda," as "it
would not make sense to someone who is skilled in the art to modify a
method of detecting local outliers as described by Pokrajac with a method to
identify problems with equipment and take countermeasures as explained in
Maeda." (Reply Br. 8; App. Br. 12.) Appellants' support for this
contention, however, relies upon Appellants' previous argument that Maeda
does not teach the correcting limitation of claim 1. (App. Br. 11-12.) As
discussed supra, we are not persuaded by Appellants' argument that Maeda
does not teach this limitation of claim 1. Additionally, the Examiner has
provided sufficient articulation for combining Maeda's technique of
correcting learning data with Pokrajac's outlier detection algorithms (also
trained on learning data) to improve identification of anomalies, thereby
11
Appeal2018-001932
Application 14/218,544
supporting the legal conclusion of obviousness. (Final Act. 5; see also
Pokrajac pp. 512-513 (describing LOP outlier detection algorithm being
trained with learning data).) See KSR Int 'l Co. v. Teleflex Inc., 550 U.S.
398,418 (2007) (quoting In re Kahn, 441 F.3d 977, 988 (Fed. Cir. 2006)).
For these reasons, we sustain the Examiner's rejection of independent
claim 1 and the Examiner's rejection of independent claims 6 and 11 on the
same basis as claim 1 (see App. Br. 5, 11-12), for the reasons stated above.
No separate arguments are presented for dependent claims 2, 3, 7, 8, 12, and
13. (App. Br. 12.) Accordingly, for the reasons stated with respect to
independent claims 1, 6, and 11, we sustain the rejection of these dependent
claims. See 37 C.F.R. § 4I.37(c)(l)(iv).
DECISION
We affirm the Examiner's decision rejecting claims 1-3, 6-8, and 11-
13 under 35 U.S.C. § 103.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l )(iv).
AFFIRMED
12