Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
SYSTEM AND METHOD FOR VALIDATING DATA
TECHNICAL FIELD
[0001] The following relates generally to validating data.
BACKGROUND
[0002] Data that is generated for or by a process, and held or used by an
organization, may be analyzed for various purposes such as to generate
statistical
reports, market insights, operational data, etc. Large quantities of
statistical data may
be generated by the organization during a period of time, e.g., on a quarterly
basis.
These large quantities of data may also need to be reviewed in a timely
manner, e.g., to
spot errors in the data and to flag or rectify those errors.
[0003] It is found that in many cases these large quantities of data are
reviewed
manually, e.g., during testing cycles. Such a manual review is time consuming
and can
be labor intensive and inefficient. These testing and review cycles may also
introduce
significant delays in identifying an issue with the source of the data, by
which time
subsequent data may have been generated with the same or similar errors.
SUMMARY
[0004] Certain example systems and methods described herein enable data
such
as statistical output data to be validated. In one aspect, there is provided a
device for
validating data. The device includes a processor, a data interface coupled to
the
processor, and a memory coupled to the processor. The memory stores computer
executable instructions that when executed by the processor cause the
processor to
obtain a validation set comprising at least one validation case, each
validation case
comprising at least one test condition. The computer executable instructions
also cause
the processor to obtain, via the communications module, at least one data set
to be
validated using the validation set. The computer executable instructions also
cause the
processor to apply the validation set to the at least one data set to validate
the data in
the data set by, for each record in the at least one data set, validating a
value in the
- 1 -
23833811.1
CA 3070681 2020-01-31
record according to the at least one test condition. The computer executable
instructions also cause the processor to output a validation result for each
record.
[0005] In another aspect, there is provided a method of validating data.
The
method is executed by a device having a data interface coupled to a processor
and
includes obtaining a validation set comprising at least one validation case,
each
validation case comprising at least one test condition. The method also
includes
obtaining, via the data interface, at least one data set to be validated using
the
validation set. The method also includes applying the validation set to the at
least one
data set to validate the data in the data set by, for each record in the at
least one data
set, validating a value in the record according to the at least one test
condition. The
method also includes outputting a validation result for each record.
[0006] In another aspect, there is provided a computer readable medium for
validating data. The computer readable medium includes computer executable
instructions for obtaining a validation set comprising at least one validation
case, each
validation case comprising at least one test condition. The computer readable
medium
also includes instructions for obtaining, via a data interface, at least one
data set to be
validated using the validation set. The computer readable medium also includes
instructions for applying the validation set to the at least one data set to
validate the
data in the data set by, for each record in the at least one data set,
validating a value in
the record according to the at least one test condition. The computer readable
medium
also includes instructions for outputting a validation result for each record.
BRIEF DESCRIPTION OF THE DRAWINGS
[0007] Embodiments will now be described with reference to the appended
drawings wherein:
[0008] FIG. 1 is a schematic diagram of an example computing environment.
[0009] FIG. 2 is a schematic diagram of another example computing
environment.
[0010] FIG. 3 is a block diagram of an example configuration of a
statistical analysis
device.
-2-
23833811.1
CA 3070681 2020-01-31
[0011] FIG. 4 is a block diagram of an example configuration of a
validation device.
[0012] FIG. 5 is a flow diagram illustrating an example of computer
executable
instructions for validating data.
[0013] FIG. 6 is a flow diagram illustrating an example of computer
executable
instructions for validating data using validation cases.
[0014] FIG. 7 is a flow diagram of an example of computer executable
instructions
for validating data using automatically generated validation cases.
[0015] FIG. 8 is a graphical illustration of a list of validation test
results with all
conditions passing.
[0016] FIG. 9 is a graphical illustration of a list of validation test
results having failed
conditions.
DETAILED DESCRIPTION
[0017] It will be appreciated that for simplicity and clarity of
illustration, where
considered appropriate, reference numerals may be repeated among the figures
to
indicate corresponding or analogous elements. In addition, numerous specific
details
are set forth in order to provide a thorough understanding of the example
embodiments
described herein. However, it will be understood by those of ordinary skill in
the art that
the example embodiments described herein may be practiced without these
specific
details. In other instances, well-known methods, procedures and components
have not
been described in detail so as not to obscure the example embodiments
described
herein. Also, the description is not to be considered as limiting the scope of
the example
embodiments described herein.
[0018] A system, devices and a process are provided to validate the results
of
statistical analyses that can eliminate at least some manual review that may
be required
on a periodic basis. The outputs of a statistical model (e.g., a scoring
model) or any
consistent data output can be analyzed using the process to flag problems and
identify
errors in the statistical results to enable an organization to investigate
failures, e.g., via
a notification, alert, or by interrupting a process that uses the results. The
methodology
-3-
23833811.1
CA 3070681 2020-01-31
described in greater detail below can also be used in a testing tool such that
as
statistical models are built and perfected, the testing tool can be used to
determine
whether the results are trending in the correct direction.
[0019] In an implementation, the details to be validated in the data may be
determined/defined in advance, prior to applying an automated validation
process.
Cases to be validated can be obtained from a possible unlimited number of
sources.
Statistical datasets to be validated may also be obtained from a possible
unlimited
number of sources. In one example, a dual looping structure may then be
applied to
validate the statistical data, with the output being a pass/fail result for
each record or
portion of the data that is analyzed.
[0020] The methodology can also be adapted for a completely automated
solution,
in which the system can automatically derive the validation cases from
existing
statistical data. In the completely automated solution, details to be
validated may be
generated based on existing data sets that are made available to the process,
which
can be validated from a possible unlimited number of sources. The dual looping
structure may also be applied to validate the statistical data and the
pass/fail results can
be output as feedback to the user, e.g., using a notification, alert, or
process instructions
such as an interruption or fault.
[0021] The automated process may therefore analyze the results of an
existing
process to validate those results. The methodology can also be applied to
incoming
data that has not yet been statistically analyzed. Machine learning may also
be used to
train the system to determine the attributes of the data to be validated, in
order to
generate and improve the automated creation of validation sets.
[0022] The process described herein can be applied to financial data (e.g.,
to
determine how much capital to set aside according to regulatory requirements),
as well
as other types of data such as medical test results, research test data, or
other
statistical data that is to be validated.
[0023] Certain example systems and methods described herein enable data
such
as statistical output data to be validated. In one aspect, there is provided a
device for
-4-
23833811.1
CA 3070681 2020-01-31
validating data. The device includes a processor, a data interface coupled to
the
processor, and a memory coupled to the processor. The memory stores computer
executable instructions that when executed by the processor cause the
processor to
obtain a validation set comprising at least one validation case, each
validation case
comprising at least one test condition. The computer executable instructions
also cause
the processor to obtain, via the communications module, at least one data set
to be
validated using the validation set. The computer executable instructions also
cause the
processor to apply the validation set to the at least one data set to validate
the data in
the data set by, for each record in the at least one data set, validating a
value in the
record according to the at least one test condition. The computer executable
instructions also cause the processor to output a validation result for each
record.
[0024] In another aspect, there is provided a method of validating data.
The
method is executed by a device having a data interface coupled to a processor
and
includes obtaining a validation set comprising at least one validation case,
each
validation case comprising at least one test condition. The method also
includes
obtaining, via the data interface, at least one data set to be validated using
the
validation set. The method also includes applying the validation set to the at
least one
data set to validate the data in the data set by, for each record in the at
least one data
set, validating a value in the record according to the at least one test
condition. The
method also includes outputting a validation result for each record.
[0025] In another aspect, there is provided a non-transitory computer
readable
medium for validating data. The computer readable medium includes computer
executable instructions for obtaining a validation set comprising at least one
validation
case, each validation case comprising at least one test condition. The
computer
readable medium also includes instructions for obtaining, via a data
interface, at least
one data set to be validated using the validation set. The computer readable
medium
also includes instructions for applying the validation set to the at least one
data set to
validate the data in the data set by, for each record in the at least one data
set,
validating a value in the record according to the at least one test condition.
The
-5-
23833811.1
CA 3070681 2020-01-31
computer readable medium also includes instructions for outputting a
validation result
for each record.
[0026] In certain example embodiments, at least one validation case can be
automatically derived by obtaining a sample data set, analyzing the sample
data set,
and identifying the at least one test condition from the analyzed sample data
set.
[0027] In certain example embodiments, at least one validation case can be
derived
by providing a user interface to enable manual entry of the at least one test
condition.
[0028] In certain example embodiments, at least one validation case can be
obtained
from a source, the source having previously derived the at least one test
condition.
[0029] In certain example embodiments, the device can obtain the sample
data set,
analyze the sample data set, and automatically identify all test conditions to
be
validated for the at least one data set to validate. The sample data set can
be analyzed
by applying an automated process that uses a model derived using a machine
learning
process.
[0030] In certain example embodiments, a notification can be generated
indicative of
at least one failure to trigger an investigation of the data set.
[0031] In certain example embodiments, validating the value in the record
can
include accessing a first record to be validated, incrementing through each of
the at
least one test condition to be validated for the first record and, for a
second and any
additional record to be validated, incrementing to a next record to increment
through
each of the at least one test condition. The validation results can include a
pass or fail
indication output as the validating increments through the values.
[0032] In certain example embodiments, the data set can be generated using
financial data. In certain example embodiments, each data set can include
statistical
results associated with use of a statistical model.
[0033] In certain example embodiments, each data set can include incoming
data to
a process.
[0034] FIG. 1 illustrates an exemplary computing environment 10 in which
data from
a data source 12 is processed, analyzed or otherwise examined by a statistical
analysis
module 14. In this exemplary environment 10, the statistical analysis module
14 is
-6-
23833811.1
CA 3070681 2020-01-31
operated by a device (not shown) having a processor, memory, and an interface
to or
with the data source 12 and obtains or receives data sets from the data source
12 via
such an interface. The statistical analysis module 14 examines the data to
perform a
statistical or other data analysis or data processing task to generate a
statistical output
16 such as a summary, report, or notification displayed in a GUI of a software
program
used by an organization or individual. The statistical output 16 can take
various forms
dependent upon the requirements or preferences of the application in which the
statistical analysis module 14 is being used.
[0035] The statistical analysis may be done for internal monitoring or
reporting for
the organization or in response to a request, audit or other internal or
external process
18 that uses the statistical output 16. For example, the process 18 may
include
generating a model scoring report that uses internal and/or external data and
is
subsequently reported to an external authority or internal body, e.g.,
analyzing credit
card balances, loans, and other customer borrowing activity to determine how
much
capital needs to be allocated to satisfy a regulatory requirement. The
statistical
analysis module 14 may be provided with, receive or otherwise obtain one or
more
statistical models 15 that define what or how to analyze the data from the
data source
12 for a particular analysis.
[0036] It can be appreciated that the computing environment 10 shown in
FIG. 1 can
be adapted to be integrated into any computing system, device, or platform,
including
an existing organization such as a financial institution. Other types of data
may also be
analyzed within a computing environment 10 such as that shown in FIG. 1, for
example,
medical testing data, research results, or any other statistical output 16 for
which the
results can be validated by testing or otherwise examining one or more test
conditions
such as ranges, upper or lower thresholds, etc.
[0037] Also shown in the computing environment 10 illustrated in FIG. 1 is
an output
validation module 20 that can be integrated or interfaced with or otherwise
coupled to
the statistical analysis module 14 in order to validate the statistical output
16. It can be
appreciated that the output validation module 20 may be operated by a device
(not
shown) having a processor, memory, and an interface to or with the statistical
output 16
-7-
23833811.1
CA 3070681 2020-01-31
and obtains or receives data sets from the statistical analysis module 14 or a
memory
device storing the statistical output 16 via such an interface. The
statistical analysis
module 14 and output validation module 20 can be hosted or provided by
separate
devices or systems or can be integrated into the same device or system. For
example,
the output validation module 20 can be provided by a separate service or
entity that can
serve multiple entities that operate or control operation of a statistical
analysis module
14 to provide a validation e.g., as a service.
[0038] The output validation module 20 obtains the statistical output 16,
e.g., as a
number of records in a data set, and analyzes the data against one or more
validation
cases 22 as discussed in greater detail below. The validation cases 22 are
obtained,
defined, or automatically determined according to one or more test conditions
26. The
test conditions 26 can be determined from or by the statistical analysis
module 14, e.g.,
based on the type of model 15, type of data, an objective of the analysis, the
expected
results, etc. The test conditions 26 can also be determined from or by
analyzing the
data source 12 directly.
[0039] The output validation module 20 can be coupled to the statistical
output
module 14 to perform a parallel validation process or, as illustrated using
dashed lines
in FIG. 1, to intercept the statistical output 16 prior to use of the output
16 by the
process 18. The output validation module 20 may therefore generate validation
results
24 that can be used to provide feedback, alerts, notifications, or control
over the
execution of the process 18. For example, the output validation module 20 can
be
inserted or coupled to a statistical analysis workflow (as illustrated in FIG.
1) to validate
the statistical output 16 periodically or in real-time as the output 16
becomes available
and prior to using or relying on the statistical output 16 in the process. In
one example
scenario, the process 18 may include generating a report for an auditor with
the
validation process being used to confirm the statistical results prior to
submitting the
report to the auditor, which may also include a feedback mechanism to have
certain
errors in data points or process operations fixed.
[0040] It can be appreciated that the computing environment 10 shown in
FIG. 1 may
also be utilized for validating the models 15 as they are being built or
trained such that
-8-
23833811.1
CA 3070681 2020-01-31
the results are analyzed and validated prior to releasing the model 15 for
production. It
can also be appreciated that the output validation module 20 can also be used
to
examine incoming data that may or may not have been statistically analyzed by
the
statistical analysis module 14. That is, the data source 12 can be directly
fed into the
output validation module 20 as illustrated in dashed lines in FIG. 1, to
enable the
contents of the data source 12 to be validated directly. For example, the data
source 12
may include externally generated data that provides what can be considered
"statistical"
values that can be directly analyzed and do not necessarily require an
additional
statistical analysis.
[0041] FIG. 2 illustrates another exemplary computing environment 30 to
which the
configuration shown in FIG. 1 has been adapted. In one aspect, the computing
environment 30 may include a statistical analysis device 36, one or more data
source
devices 32 providing or otherwise having access to external data sources 12a,
and a
communications network 34 connecting one or more components of the computing
environment 30. The computing environment 30 may also include one or more
validation devices 38. In the example shown in FIG. 2, the validation device
38 receives
data via the statistical analysis device 36 after the data has undergone a
statistical
analysis by a statistical analysis module 14, but can also access an external
data
source 12a via the communications network 34 to directly validate one more
data sets
from the external data source 12a. In one example, the validation device 38
may be
associated with another organization that relies on the data after having been
processed by the statistical analysis device 36. The devices 36, 38 may also
be part of
the same organization, and/or may be integrated into a single device (not
shown).
[0042] The computing environment 30 may also include one or more 3rd party
devices 40. The 3rd party device 40 may be considered similar to the devices
36, 38 but
in this example does not necessarily process or analyze the data. For example,
the 3'd
party device 40 may correspond to a member of the public that consumes a
report,
score, or result generated by the process 18, or may correspond to an auditor
or other
external organization that relies on the statistical output 16.
-9-
23833811.1
CA 3070681 2020-01-31
[0043] It can be appreciated that the 3rd party device 40 may also receive
data that
has been validated by the validation device 38 (as illustrated in dashed lines
in FIG. 2).
It can also be appreciated that the validation device 38 and 3rd party device
40 may
include an application programming interface (API) or other interface
mechanism or
module for interfacing with the statistical analysis device 36 (or each other)
either
directly or via the network 34. Similarly, the statistical analysis device 36
may include
an API or other interface mechanism or module for interfacing with the
external data
source 12a via the data source device 32. The data source device 32 is shown
to
illustrate one example in which an entity or organization responsible for the
external
data source 12a communicates with the statistical analysis device 36 and/or
the
validation device 38 via the network 34. However, in other configurations, the
statistical
analysis device 36 and/or validation device 38 may be capable of accessing the
external data source 12a directly, without communicating via another device.
It can be
appreciated that a statistical analysis device 36 may in another scenario
become a
validation device 38 and vice versa. As such, the scenario and configuration
depicted in
FIG. 2 provides one example for the sake of illustration.
[0044] As illustrated in FIG. 2, the statistical analysis device 36 may
also include or
have access to an internal data source 12b, that is, data that is generated or
otherwise
made available within a same entity or organization. For example, data
generated in
one business unit of a financial institution may be used in other downstream
processes
18 and therefore could benefit from execution of the statistical analysis
module 14 prior
to using the internally sourced data 12b. In one embodiment, the statistical
analysis
device 36 may be one or more computer systems configured to process and store
information and execute software instructions to perform one or more processes
consistent with the disclosed embodiments. Similarly, the validation device 38
can
have, or have access to, the internal data source 12b shown in FIG. 2 or its
own internal
data source 12b (not shown).
[0045] The statistical analysis device 36 and/or validation device 38 may
also
include or be a component or service provided by a financial institution
system (e.g.,
commercial bank) that provides financial services accounts to users, processes
financial
-10-
23833811.1
CA 3070681 2020-01-31
transactions associated with those financial service accounts, and analyzes
statistical
data to inform investors, customers, and the public generally. Details of such
a financial
institution system have been omitted for clarity of illustration. The
statistical analysis
device 36 and/or validation device 38 may also include or be a component or
service
provided by other types of entities and organizations, such as government
bodies and
private enterprises that would benefit from checking the integrity of data
which they did
not necessarily generate.
[0046] In certain aspects, data source device 32 (that provides or provides
access to
the external source of data 12a), statistical analysis device 36, and/or
validation device
38 can include, but is not limited to, a personal computer, a laptop computer,
a tablet
computer, a notebook computer, a hand-held computer, a personal digital
assistant, a
mobile phone, an embedded device, a smart phone, a virtual reality device, an
augmented reality device, third party portals, and any additional or alternate
computing
device, and may be operable to transmit and receive data across communication
network 34.
[0047] Communication network 34 may include a telephone network, cellular,
and/or
data communication network to connect different types of devices as will be
described
in greater detail below. For example, the communication network 34 may include
a
private or public switched telephone network (PSTN), mobile network (e.g.,
code
division multiple access (CDMA) network, global system for mobile
communications
(GSM) network, and/or any 3G, 4G, or 5G wireless carrier network, etc.), WiFi
or other
similar wireless network, and a private and/or public wide area network (e.g.,
the
Internet).
[0048] The computing environment 30 may also include a cryptographic server
(not
shown) for performing cryptographic operations and providing cryptographic
services
(e.g., authentication (via digital signatures), data protection (via
encryption), etc.) to
provide a secure interaction channel and interaction session, etc. Such a
cryptographic
server can also be configured to communicate and operate with a cryptographic
infrastructure, such as a public key infrastructure (PKI), certificate
authority (CA),
-11-
23833811.1
CA 3070681 2020-01-31
certificate revocation service, signing authority, key server, etc. The
cryptographic
server and cryptographic infrastructure can be used to protect the various
data
communications described herein, to secure communication channels therefor,
authenticate parties, manage digital certificates for such parties, manage
keys (e.g.,
public and private keys in a PKI), and perform other cryptographic operations
that are
required or desired for particular applications of the statistical analysis
device 36,
validation device 38, 3rd party device 40, and data source device 32. The
cryptographic
server may be used to protect the data or results of the data by way of
encryption for
data protection, digital signatures or message digests for data integrity, and
by using
digital certificates to authenticate the identity of the users and devices
within the
computing environment 30, to inhibit data breaches by adversaries. It can be
appreciated that various cryptographic mechanisms and protocols can be chosen
and
implemented to suit the constraints and requirements of the particular
deployment of the
computing environment 30 as is known in the art.
[0049] In FIG. 3, an example configuration of the statistical analysis
device 36 is
shown and in FIG. 4, an example configuration of the validation device 38 is
shown. As
can be appreciated from these figures, the devices 36, 38 include several
similar
components, which will be described once for brevity. Turning first to FIG. 3,
in certain
embodiments, the statistical analysis device 36 may include one or more
processors 50,
a communications module 52, and a data interface module 54 for interfacing
with the
external data source 12a and/or internal data source 12b to retrieve and store
data.
Communications module 52 enables the statistical analysis device 36 to
communicate
with one or more other components of the computing environment 30, such as
data
source device 32, validation device 38, 3rd party device 40 (or one of its
components),
via a bus or other communication network, such as the communication network
34.
While not delineated in FIG. 3, the statistical analysis device 36 includes at
least one
memory or memory device that can include a tangible and non-transitory
computer-
readable medium having stored therein computer programs, sets of instructions,
code,
or data to be executed by processor 50. FIG. 3 illustrates examples of
modules, tools
and engines stored in memory on the statistical analysis device 36 and
operated by the
- 12 -
23833811.1
CA 3070681 2020-01-31
processor 50. It can be appreciated that any of the modules, tools, and
engines shown
in FIG. 3 may also be hosted externally and be available to the statistical
analysis
device 36, e.g., via the communications module 52.
[0050] In the example embodiments shown in FIGS. 3 and 4, the statistical
analysis
device 36 and validation device 38 include a machine learning engine 56, a
classification module 58, a training module 60, an output module 64, and a
process
interface module 66. Referring to FIG. 3, the statistical analysis device 36
also includes
the statistical analysis module 14 storing or having access to one or more
statistical
models 15. Similarly, as shown in FIG. 4, the validation device 38 also
includes the
output validation module 20 and may store one or more validation cases 22.
[0051] The machine learning engine 56 is used by the statistical analysis
module 14
or output validation module 20 to generate and train statistical models 15 or
validation
cases 22 to be used in either the statistical analyses being conducted,
building or
refining the models 15, determining validation cases 22, and performing a data
validation process. The statistical analysis module 14 or output validation
module 20
may utilize or otherwise interface with the machine learning engine 56 to both
classify
data currently being analyzed to generate the statistical models 15 or
validation cases
22, and to train classifiers using data that is continually being processed
and
accumulated by the statistical analysis device 36 and validation device 38.
[0052] The machine learning engine 56 may also perform operations that
classify the
data from the data source(s) 12a/12b in accordance with corresponding
classifications
parameters, e.g., based on an application of one or more machine learning
algorithms
to the data. The machine learning algorithms may include, but are not limited
to, a one-
dimensional, convolutional neural network model (e.g., implemented using a
corresponding neural network library, such as Keras ), and the one or more
machine
learning algorithms may be trained against, and adaptively improved using,
elements of
previously classified profile content identifying expected datapoints.
Subsequent to
classifying the data, the machine learning engine 56 may further process each
data
point to identify, and extract, a value characterizing the corresponding one
of the
- 13 -
23833811.1
CA 3070681 2020-01-31
classification parameters, e.g., based on an application of one or more
additional
machine learning algorithms to each of the data points. By way of the example,
the
additional machine learning algorithms may include, but are not limited to, an
adaptive
natural language processing algorithm that, among other things, predicts
starting and
ending indices of a candidate parameter value within each data point, extracts
the
candidate parameter value in accordance with the predicted indices, and
computes a
confidence score for the candidate parameter value that reflects a probability
that the
candidate parameter value accurately represents the corresponding
classification
parameter. As described herein, the one or more additional machine learning
algorithms
may be trained against, and adaptively improved using, the locally maintained
elements
of previously classified data. Classification parameters may be stored and
maintained
using the classification module 58, and training data may be stored and
maintained
using the training module 60.
[0053] In some instances, classification data stored in the classification
module 58
may identify one or more parameters, e.g., "classification" parameters, that
facilitate a
classification of corresponding elements or groups of recognized data points
based on
any of the exemplary machine learning algorithms or processes described
herein. The
one or more classification parameters may correspond to parameters that can
identify
expected and unexpected data points for certain types of data.
[0054] In some instances, the additional, or alternate, machine learning
algorithms
may include one or more adaptive, natural-language processing algorithms
capable of
parsing each of the classified portions of the data being examined and
predicting a
starting and ending index of the candidate parameter value within each of the
classified
portions. Examples of the adaptive, natural-language processing algorithms
include, but
are not limited to, natural-language processing models that leverage machine
learning
processes or artificial neural network processes, such as a named entity
recognition
model implemented using a SpaCy library.
[0055] Examples of these adaptive, machine learning processes include, but
are not
limited to, one or more artificial, neural network models, such as a one-
dimensional,
- 14 -
23833811.1
CA 3070681 2020-01-31
convolutional neural network model, e.g., implemented using a corresponding
neural
network library, such as Keras . In some instances, the one-dimensional,
convolutional
neural network model may implement one or more classifier functions or
processes,
such a Softmax classifier, capable of predicting an association between a
data point
and a single classification parameter and additionally, or alternatively,
multiple
classification parameters.
[0056] Based on the output of the one or more machine learning algorithms
or
processes, such as the one-dimensional, convolutional neural network model
described
herein, machine learning engine 56 may perform operations that classify each
of the
discrete elements of the data being examined as a corresponding one of the
classification parameters, e.g., as obtained from classification data stored
by the
classification module 58.
[0057] The outputs of the machine learning algorithms or processes may then
be
used by the statistical analysis module 14 to generate and train the models 15
and to
use the models 15 to determine if data points in the current data being
examined are
expected or unexpected. The outputs of the machine learning algorithms or
processes
may also be used by the output validation module 20 to generate and train
validation
cases 22 to determine if data points in the current data being examined are
expected or
unexpected.
[0058] Referring again to FIGS. 3 and 4, the output module 64 may be used
to
provide one or more outputs based on the results generated by the statistical
analysis
module 14 or output validation module 20. Example outputs include a visual
output in a
GUI; a flag, alert, notification, or message in a process using (or about to
use) the data
being analyzed and/or validated; or a process instruction operable to pause,
interrupt or
halt the process in view of the results of the data validation as discussed
above. The
output module 64 may be configured to interface with the process 18 via the
process
interface module 66. The statistical analysis module 14 and output validation
module
20 may also be configured to interface with the process 18 via the process
interface
module 66. The output module 64 and process interface module 66 may be
embodied
- 15 -
23833811.1
CA 3070681 2020-01-31
as APIs when interfacing with software-based processes 18 or may include a
combination of software and hardware when interfacing with processes 18 that
have
hardwired or software/hardware-type interfaces. The statistical analysis
module 14 and
output validation module 20 may be programmed to translate between multiple
protocols in order to interface with other components to provide such outputs
and such
translation can occur within the statistical analysis module 14, output data
validation
module 20 and/or the output module 64 or process interface module 66. It can
be
appreciated that the functionality provided by the output module 64 and
process
interface module 66 are delineated as shown in FIGS. 3 and 4 for illustrative
purposes
and such functionality may also be integrated together or into the statistical
analysis
module 14 or output data validation module 20 in other example embodiments.
[0059] While not shown in the figures, the 3rd party device 40 may also be
configured
in a manner similar to the devices 36, 38 to enable the 3rd party device 40 to
report,
publish, or otherwise use the data from a data source 12 that has been
processed by
either or both the devices 36, 38.
[0060] It will be appreciated that any module or component exemplified
herein that
executes instructions may include or otherwise have access to computer
readable
media such as storage media, computer storage media, or data storage devices
(removable and/or non-removable) such as, for example, magnetic disks, optical
disks,
or tape. Computer storage media may include volatile and non-volatile,
removable and
non-removable media implemented in any method or technology for storage of
information, such as computer readable instructions, data structures, program
modules,
or other data. Examples of computer storage media include RAM, ROM, EEPROM,
flash memory or other memory technology, CD-ROM, digital versatile disks (DVD)
or
other optical storage, magnetic cassettes, magnetic tape, magnetic disk
storage or
other magnetic storage devices, or any other medium which can be used to store
the
desired information and which can be accessed by an application, module, or
both. Any
such computer storage media may be part of the data source device 32,
statistical
analysis device 36, validation device 38, or 3rd party device 40, or
accessible or
- 16 -
23833811.1
CA 3070681 2020-01-31
connectable thereto. Any application or module herein described may be
implemented
using computer readable/executable instructions that may be stored or
otherwise held
by such computer readable media.
[0061] Referring to FIG. 5, an example illustrating computer executable
operations
executed by the output validation module 20 in performing a data validation
process is
shown, for example in validating data such as statistical output 16 generated
by the
statistical analysis module 14. At block 80, the output validation module 20
obtains a
validation set with one or more validation cases 22. Each validation case 22
includes at
least one test condition 26 (or criteria to validate ¨ see FIGS. 7 and 8),
which defines
what in the data is to be validated. For example, if a value in the data set
should be
within a predetermined range in order to be valid, the test condition 26 can
define that
range and the validation case 22 apply that condition. The terms validation
set,
validation case, and test condition are used in this granularity to provide
for flexible and
reusable modules, however, a validation set may instead define one or more
test
conditions or the test condition may itself be the validation case or
validation set. On
the other hand, different validation cases can be defined with different test
conditions
that can be assembled in different ways for different types of data sets
associated with
different types of statistical analyses and/or statistical models 15. For
example, Test
Case A may include test condition 1 and test condition 2 while Test Case B
includes
test condition 1 and test condition 3. Test Case A may be selected for data
from one
region while Test Case B is selected for another region. In this way,
different test cases
22 can be defined, reused and applied modularly.
[0062] At block 82, the output validation module 20 obtains the one or more
data
sets to be validated using the validation set. This may include, for example,
communicating with the statistical analysis device 36 via the process
interface module
66 of the validation device 38 to obtain the statistical output 16 generated
by the
statistical analysis module 14 and which is to be validated.
[0063] At block 84, the output validation module 20 applies the validation
set to the
one or more data sets that are being validated, to validate the data in the
one or more
-17-
23833811.1
CA 3070681 2020-01-31
data sets according to the test conditions 26. The validation set can be
applied to the
data being validated by iterating through data fields, rows, columns or other
records
according to what is being validated. An example of the application of a
validation set is
described in greater detail below.
[0064] At block 86, the output validation module 20 outputs a validation
result 24 for
each record that is validated, with an indication of whether the value that
was examined
passed or failed according to the test condition(s) 26. For example, if the
value
contained in a record is outside of a range that is expected given some other
condition
or variable, the validation result 24 for that record would fail. The
outputting of the
validation result at block 86 may include generating a summary or list of the
records
with the corresponding result, e.g., within a user interface.
[0065] At block 88, the validation results 24 may optionally be provided to
a process
18 that uses the statistical output 16 as a validation feedback mechanism. For
example, the output validation module 20 may be initiated and called by the
statistical
analysis module 14 or statistical analysis device 36 to perform a scheduled
validation or
to validate results in real-time before executing a further action in the
process 18.
[0066] FIG. 6 provides an example implementation of the operations shown in
FIG.
5. In this example embodiment, an existing process is running at block 100.
This
existing process can include any process, such as an execution of the
statistical
analysis module 14 or the process 18, that has data to be validated by the
output
validation module 20. At block 102 an option is provided for running the
validation
process. When this option is selected, a number of validation cases 22 are
obtained at
stage 104. In this example, the validation cases 22 are shown as a series of
files 1,2,
..., n; illustrating that any number of validation cases 22 can be used to
perform the
validation process. At block 106 a validation set is built for processing.
This may
include assembling a list of test conditions 26 to be applied to each record
or other
portion of data being validated.
[0067] At stage 108, a number of data sets to be validated is defined, with
each
being evaluated against one or more validation cases 22. In this example, a
series of
- 18 -
23833811.1
CA 3070681 2020-01-31
data sets 1, 2, ..., n is shown; illustrating that any number of data sets 12,
16 can be
obtained for validation in stage 108. It may be noted that the process shown
in FIG. 6
can be applied to statistical output 16 or incoming data 12 as discussed
above. The
data sets 12, 16 to be validated in stage 108 may be evaluated individually at
block 110
as shown in FIG. 6 or may be combined and evaluated together depending on the
test
conditions 26 being applied and the data being analyzed.
[0068] At block 112, the output validation module 20 initiates a dual
looping
structure, which is an example implementation for block 84 shown in FIG. 5.
Here the
first (or next) record in the data set being validated is accessed. At block
114, the
record(s) of the conditions to validate is/are obtained. This may include
determining all
of the test conditions 26 to apply against the data record currently being
analyzed, as
defined by the validation set. At block 116 the output validation module 20
determines
if the data record is valid for the current test condition 26 being analyzed.
If not, the
output validation module 20 outputs a failure at block 118. If the data record
is valid for
the current test condition 26, the output validation module 20 determines at
block 120 if
the current test condition 26 is the final test condition 26 to validate for
that data record.
If not, the output validation module 20 increments to the next test condition
26 at block
122 and repeats blocks 114-120 until it is determined at block 120 that all
criteria have
been analyzed for validity.
[0069] When all criteria have been analyzed for validity, the output
validation module
20 determines at block 124 if the current data record is the final data record
to be
analyzed. If not, the output validation module 20 increments to the next data
record and
repeats blocks 112-124 for the next data record. Once all data records have
been
analyzed, at block 128 the output validation module 20 outputs a pass result
for the data
set and returns to the existing process at block 130. It can be appreciated
that the dual
looping structure shown in FIG. 6 can be applied in parallel or serially to
the other data
sets to be validated that are shown in stage 108. In this way, the dual
looping structure
can be applied to a data set to determine if any of the data records of that
type contain
an error wherein the failure output at block 118 is indicated.
- 19 -
23833811.1
CA 3070681 2020-01-31
[0070] It can also be appreciated that the validation cases 22 determined
in stage
104 may be predetermined, selected, or otherwise specified at the time of, or
prior to,
running the validation process at block 102. Such predetermined validation
cases 22
may be specified by an owner of the data source 12, an operator of the
statistical
analysis module 14, a third party such as a regulatory or government body, or
any other
interested party.
[0071] FIG. 7 provides another example embodiment for validating data using
the
output validation module 20, in a fully automated configuration. In this
example
embodiment, the fully automated validation process may be selected at block
200 for
data associated with the existing process 100. In this example embodiment, the
validation process 200 is executed to automatically determine the validation
set to be
applied to the data sets to be validated. When the validation process is
initiated at block
200, a data set used to build the validation set is obtained at block 202. The
data set is
analyzed at block 204 to determine which validation cases 22, including which
test
conditions 26, are to be applied to the rest of the data. For example, the
analyses
conducted at block 204 can include applying a machine learning algorithm to
the data
set to determine test conditions 26 indicative of whether a data record is
valid or not.
The machine learning algorithm can be trained based on validation sets
determined
using the process shown in FIG. 6 or by applying a model based on the type of
data
being validated.
[0072] Once the validation set is built for processing at block 206, the
data sets to be
validated are determined at stage 108, and the dual looping structure
described above
can be applied beginning at block 110. The implementation of blocks 110-130
are
described above and need not be reiterated here.
[0073] An example of an output page 300 is shown in FIG. 8. In the example
page
300, a validation output 304 is displayed for each validation case in a list
of results 302.
It can be seen that in this example all test conditions 26 passed the
validation process.
- 20 -
23833811.1
CA 3070681 2020-01-31
[0074] Another example of an output page 400 is shown in FIG. 9. In the
example
page 400, a validation output 306 is shown that corresponds to a failed test
condition
306.
[0075] It will be appreciated that the examples and corresponding diagrams
used
herein are for illustrative purposes only. Different configurations and
terminology can be
used without departing from the principles expressed herein. For instance,
components
and modules can be added, deleted, modified, or arranged with differing
connections
without departing from these principles.
[0076] The steps or operations in the flow charts and diagrams described
herein are
just for example. There may be many variations to these steps or operations
without
departing from the principles discussed above. For instance, the steps may be
performed in a differing order, or steps may be added, deleted, or modified.
[0077] Although the above principles have been described with reference to
certain
specific examples, various modifications thereof will be apparent to those
skilled in the
art as outlined in the appended claims.
- 21 -
23833811.1
CA 3070681 2020-01-31
