System And Method For Pattern-Recognition Based Monitoring and
Controlled Processing of Data Objects Based on Conformity
Measurements
Field of the Invention
The present invention relates to a pattern-recognition-based control and
monitoring system for automated underwriting systems or automated patternrecognition
based data-capturing systems providing risk sharing of risk exposed
components. In particular, the present invention relates to control systems for
monitoring the workflow of the underwriting system or of pooled underwriting objects by
capturing workflow objects in the underwriting process flow, and/or underwriting
objects from a source of pooled objects, wherein additional process states are
triggered, when non-conforming workflow objects are detected within the process
flow, by means of the control and monitoring system. More particularly, the invention
also relates to a control system for monitoring and dynamically adapting portfolios of
pooled risk transfer objects.
Background of the Invention
In consideration of the wide deployment of risk transfer technologies
worldwide, in particular a s provided by insurance and reinsurance systems, the
importance of these systems in maintaining the operation and operational conditions
on the industry-wide level becomes apparent. However, a technological approach to
many problems in this field is often difficult. For example, it is important to improve and
adjust the physical design and technological implementation of such risk transfer
systems in order to cope with any emerging problems; problems involving, e.g., the fact
that the risk assessment can be based on reproducible results, that the error rate of the
measurement of pooled risks can be minimized, that the enormous amount of data
can be processed and taken into account in the measurements, and that the
operation of such systems can be adjusted to quickly changing environmental
condition-related parameters, and/or improved by self-adaptation. Furthermore, it is
also important, that the use of the enormous resources of data can be systematized
and dynamically used, which requires appropriate technical modalities and physical
system designs.
Significant concerns for most automated risk transfer systems are data
quality and data quantity. Estimates in the industry have found that about 25% of the
operational times of insurance systems are expended on data quality issues. Moreover,
related to the technical impact of bad data quality, a further survey found that about
30% of automation problems are due to poor data quality and that automated
analyses are adversely affected by data quality issues, thereby often rendering
automated risk transfer systems too unreliable for self-sufficient day-to-day live
operations, thus reducing them to stand-alone running systems; i.e., systems that
operate without human interaction and control, thereby measuring and reacting in a
self-adapting manner to on changing environmental conditions. Some basic
technological problems are merely related to the accuracy of data and the amount
and complexity of data to be processed. Some other technical issues relate to
characteristics, such a s completeness and timeliness; however, these issues often end
again in the problem of data recognition, data acquisition, data quality and, finally
data processing. The measurement and determination of risks, in particular during the
process of risk pooling of exposed components by means of resource pooling systems
and, more particularly, a t the level of fixing the individual risk transfer condition
parameters, are technical key features in the operation of automated risk-transfer and
risk-pooling systems. However, measurements of the risk, a s it relates to a specific risk
transfer, requires the systems to operate based o n probable most up-to-date data,
which, in particular, requires not only a modality for measuring and capturing
appropriate measuring parameters but also appropriately fast and reliable data
recognition and processing. It must be added to the above comments that cheap
data storage, along with changes in regulatory requirements, have led to extraordinary
amounts of data being captured, stored, and provided to insurance systems. On one
hand, the processing of these data quantities requires appropriately adapted systems,
a s mentioned. One the other hand, these enormous amounts of data also amass a n
enormous error incidence and inconsistencies, which hinders the operation of
automated systems. Therefore, on the other hand, it is important to ensure that, already
on the technical level of generating new data, monitoring modalities are in place by
means of appropriately implemented and fast-reacting control and monitoring systems.
For automated underwriting systems providing the basis for the automated
adaption and determination of condition-related parameters of associated risk
transfers, an underwriting process-flow for underwriting workflow objects comprises the
technical and/or procedural steps required for executing the underwriting process with
regard to a n object; i.e., the underwriting process, the technical and other means to
conduct the processing steps, and the transfer and flow of data/signaling between the
means and/or steps for executing the process on the object. Each step is defined by a
set of processes, activities or tasks that need to be implemented. Within a n underwriting
workflow, objects for underwriting (e.g., risk transfer objects comprising operational
condition parameters for the risk transfer, i.e. technical objects affecting the operation
and interaction of the resource pooling system and providing risk cover by pooling
resources from risk exposed components, and triggering risk events in order to
automatically cover the impact of occurring risk events by means of the transfer of the
pooled resources) pass through the different steps in the specified order, from start to
finish, and the underwriting processes at each step are executed either by dedicated
technical processing devices or means, by activating specific system functionalities
(also, e.g., computer program products), or by dedicated signaling to specific devices
or people intended to perform activities on the object. Automated underwriting
workflow systems can be set up using a visual front end, or they can be hard-coded,
and their execution is delegated to a workflow execution engine that handles the callup
and signal generation of the remote devices or applications.
In the prior art, underwriting workflow systems for automating risk transfer
belong to the field of so-called production workflow systems. The production or
industrial process systems are dedicated to steering and executing processing steps of
technical objects, such a s operational parameters, devices or products, by steering
and operating appropriate devices for executing the activities of the workflow objects.
Regarding data objects, the process systems also serve for functional processing and
the computation of data objects, in particular for the purpose of standardizing the
operational interaction of such systems by generating and adjusting workflow objects
by means of processing them within the workflow, i.e. the underwriting process flow.
Concerning the monitoring and control systems for such automated
underwriting systems, the prior art envisions automated underwriting workflow systems
that are able to provide various capabilities for the monitoring of workflow processes,
which are modeled and executed within the workflow system. Such capabilities can
include, for example, analysis tools for the measurement and display of metrics with
respect to the status of the processes, times for executing work steps in the context of
the processes, and [management of] bottlenecks within the processes. These
capabilities can also be transferred to the underwriting workflow system for underwriting
workflow processes, which are executed in systems that are external to the underwriting
workflow system. Many underwriting systems of the prior art comprise, a s the core, a
workflow execution engine, a process management system or a similar control
device/system for controlling and monitoring the processing of the workflow objects.
The workflow execution engine of the workflow systems can, e.g., be implemented a s a
processor-based automation means of the underwriting process flow. The workflow
execution engine steers a sequence of activities (process tasks), interactions and
signaling with execution devices or means, or in interaction with human resources
(users) or IT resources (software applications and databases), a s well a s rules for
controlling the progression of processes throughout the various stages that are
associated with each activity.
However, a t the various stages of the underwriting process, activities
typically require human interactions: i.e., user control or data entry through a form. For
certain underwriting workflow systems, one of the ways for automating and operating
the steering and monitoring tasks of the processes by means of a workflow execution
engine is to develop the appropriate processor codes and applications that guide a
processor-based workflow execution engine for the execution of the required steps of
the underwriting process; however, in practice, such underwriting workflow execution
engines are not able to accurately execute all the steps of the underwriting process by
means of the underwriting workflow system while assuring, thereby, the operational
stability of the underlying resource pooling system for the pooled risks. To solve this
problem, in the prior art, the typical approach envisions the use of a combination of
software and human intervention; however, this approach is quite complex rendering
the reproducibility, the predictability, and even the information flow and
documentation process difficult. Further, with this approach, due the amount of
processed workflow objects, it is impossible to provide a n automated control for the
underwriting systems that operate dynamically and optimized based on a monitoring of
all processed underwriting workflow objects.
Another problem in the prior art underwriting system is that workflows are
difficult to generate and/or to adapt dynamically, due to a lack of the appropriate
measuring, monitoring and control systems, which allow for filtering and dynamically
recognizing non-conforming workflow objects. Moreover, upon reaching a certain
process step in the workflow, it can become necessary to make adjustments to the
processing by way of steps which are not predictable at the beginning of the
underwriting process flow or workflow, and which can depend on environmental
parameters or operational parameters of the risk exposed components and/or the
resource pooling systems, i.e. the automated insurance systems. However, such a n
adaptation of the underwriting conditions and/or the pooled risk portfolio of transferred
risk critically depend on correctly measuring and recognizing non-conforming
underwriting objects, and, even more so, on a correct measurement of the possible
impact of such non-conforming underwriting objects based on their level of no n
conformity.
Summary of the Invention
It is one object of the present invention to provide a system and method for
a pattern-recognition-based control and monitoring system for underwriting systems or
other pattern-recognition based automated data capturing systems, which provides
risk sharing for risk exposed components. The systems can, in particular, be used in
monitoring large data transfers, controlling automated risk underwriting, a s automated
insurance systems with appropriate interfaces, or automated surveillance of networks,
especially the worldwide backbone network. Envisioned, in particular, are means for
controlling systems for monitoring and filtering of non-conforming risk objects, a s e.g .
underwriting risk-transfer objects, and also for monitoring the workflow of the
underwriting system by capturing workflow objects in the underwriting process flow,
which do not have the drawbacks of the prior art. In particular, it is an object of the
present invention to provide a monitoring and control system that is better capable of
capturing the external and/or internal factors that may affect the pooled risk of the
automated insurance system and that can further affect the processing of a n object
within a n automated underwriting workflow. Furthermore, it is an object of the present
invention to provide a system that is capable of being operated by monitoring and
recognizing non-conforming objects, thereby reacting to externally or internally
occurring boundary conditions or constraints. Furthermore, it is an object of the present
ό
invention to provide a system that is able to dynamically react to changing
environmental or internal conditions or measuring parameters, which are possibly not
known or predictable during the process of risk pooling; i.e., the underwriting process, in
particular, for recognizing, measuring and classifying automatically, i.e. monitoring nonconforming
risk-transfer underwriting objects based on their level of non-conformity.
According to the present invention, these objects are achieved particularly
by the features of the independent claims. In addition, further advantageous
embodiments can be derived from the dependent claims and their description.
According to the present invention, the above-mentioned objects for
controlling and monitoring insurance underwriting systems with regard to risk exposed
components are achieved, in particular, by controlling and monitoring the underwriting
workflow of the underwriting system and/or the portfolio of pooled risk-transfer objects,
wherein workflow objects are captured in the underwriting process flow and/or in a
source of pooled risk-transfer objects, and wherein a n activation of additional process
states is triggered upon the detection of non-conforming workflow objects within the
process flow and/or pooled risk-transfer objects of the portfolio source by means of the
control and monitoring system, and in that a plurality of block elements is generated
based on predefined boundary conditions provided by a n automated underwriting
system, wherein the block elements are triggerable parts of workflow objects and
comprise one or more of interconnectable search terms and/or meta-data, and
wherein the block elements are assigned to a recognition block-map and stored by
means of a first searchable data structure provided by a first database, and in that the
system comprises a second database providing a second searchable data structure for
storing a plurality of workflow objects, and wherein the stored workflow objects of the
second database are at least partly generated based on definable boundary
condition parameters and/or based on historical workflow objects, and in that a
workflow object is captured within the workflow pathway by means of the
measurement and monitoring system, and wherein the system comprises a core engine
with a recognition module for scanning the captured workflow object and thereby
recognizing and identifying block elements of the stored recognition block-map in the
captured workflow object, and in that, based on the filtered block elements of the
captured workflow object, a proximity factor relative to each of the stored workflow
objects of the second database is determined, wherein a corresponding proximity
factor is measured by matching recognized block elements of the captured workflow
object with block elements of a workflow object of the second database, thereby
providing the measure for the proximity of the two workflow objects based on the
mutually allocatable block elements, and in that a conformity index is generated and
assigned to the captured workflow object based on the conformity of the recognized
block elements with the stored block elements of the first searchable data structure,
and based on the conformity of the recognized block elements with workflow objects
of the second database, and wherein only workflow objects within a predefined range
of the proximity factor are selected for generating the conformity index, and wherein
the conformity index provides a measure for the conformity of the captured workflow
object with the predefined boundary conditions of the automated underwriting system,
and in that the additional process states are triggered and assigned to the process flow
by means of a process management engine of the measurement and monitoring
system when the measured conformity index exceeds a risk threshold value by means
of the trigger module. For the purpose of recognizing and identifying block elements by
means of the recognition module, the captured workflow object can, e.g., be contentscanned
by means of the recognition module, triggering block elements of the
workflow objects by means of the stored block elements of the first searchable data
structure, and wherein triggered block elements of the captured workflow object are
classified by means of a classification module, and wherein associated entities are
extracted by means of an entity extraction module, and wherein detected search
terms of the workflow object are flagged by means of a key-word flagging module. The
plurality of block elements can, e.g., be generated based on predefined boundary
conditions provided by a n automated underwriting system, generated by at least
partially extracting data from historical and/or simulated data. The one or more of the
interconnectable search terms and/or meta-data can, e.g., be extracted and/or
generated at least partially from historical and/or simulated data. The stored workflow
objects of the second database can, for example, be at least partially generated
based on definable boundary condition parameters, selecting block elements from the
first database and/or extracting workflow objects from historical workflow objects' data.
The corresponding proximity factor can, e.g., be measured by matching recognized
block elements of the captured workflow object with block elements of a workflow
object of the second database, thereby providing the measure for the proximity of the
two workflow objects based on the mutually allocatable block element s, and/or their
degree of conformity. As a variant, based on the filtered block elements of the
captured workflow object, a proximity factor with regard to each of the stored
workflow objects of the second database can, e.g., be determined, wherein a
corresponding proximity factor is measured by matching recognized block elements of
the captured workflow object with block elements of a workflow object of the second
database, thereby providing the measure for the proximity of the two workflow objects
based on the number of mutually allocatable block elements, and/or their degree of
conformity. As a further variant, the conformity index can, e.g., be generated and
assigned to the captured workflow object based on the conformity of the recognized
block elements relative to the stored block elements of the first searchable data
structure and based on the conformity of the recognized block elements with workflow
objects of the second database, and wherein the conformity index provides a measure
for the variance of the captured workflow objects with the predefined boundary
conditions of the automated underwriting system. As another variant, the control
system and the runtime execution devices can, e.g., interact in runtime, wherein the
underwriting workflow object can be processed, based on a dynamically adapted
process flow, with the generated process tasks by executing the activated process
tasks by means of the runtime execution devices based on the transmitted control and
steering signaling. A first-tier execution of the control system can, e.g., comprise the
generation of the process states, and a second-tier execution of the control system
can, e.g., comprise the generation and association of the additional process states
and/or tasks, upon measuring the conformity index exceeding the risk threshold value
by means of the trigger module, and wherein the processing of the workflow object
and the appropriate signaling by the control system is dynamically adapted by
alternating operating parameters of the runtime execution devices. Furthermore, the
capturing means can, e.g., comprise measuring devices and/or measuring sensors for
capturing the state parameters and/or task parameters of the workflow object. The
workflow objects can, e.g., be captured via a plurality of input devices that are
accessible by risk exposed components and/or automated insurance systems for
dynamically adapting the workflow objects via the interface module and/or network
based on appropriate signal generation by means of the signaling module of the
control system. At least parts of task parameters of the process tasks of the additional
process states can, for example, be dynamically generated by the control system
based on the measured conformity index of the captured workflow object. Appropriate
signaling can, e.g., be generated by means of the signaling module for steering the
execution devices that process the selected object according to the process flow
and/or additional process flow generated by the control system. The processing of the
captured workflow object can, e.g., be monitored by means of dedicated monitoring
and/or measuring devices of the control system based o n at least the captured state
parameters, the tasks parameters and/or operating parameters. The invention has, inter
alia, the advantage that a n underwriting workflow object can be processed in a n
automated process flow, wherein the process flow can be fully controlled and
operated by the monitoring and control system. Furthermore, a dynamic assignment of
operating parameters allows for the dynamic adaptation of the underwriting process
during the processing of a captured object; i.e., a n object processed in the
underwriting process flow. This also has the advantage that the control system, which is
implemented a s a dynamically adaptable system, can be automatically optimized
without any additional technical or human intervention. The present monitoring and
control system for state-transition-controlled processing of objects for responsive process
management allows operators to have real-time visibility of their processes (executing
them both within and externally of the platform), for the purpose of monitoring and
dynamically adapting their underwriting processes, executing these processes by
execution-type devices and appropriate signaling to those devices, sensing and
responding to external events, and by incrementally improving these processes. This is
not possible with the systems that are provided by the prior art. Furthermore, the
monitoring and control system allows a structuring of facultative and contract covers
(risk coverage, wordings, terms, and conditions) by means of the appropriately
adjusted underwriting objects, identification of capacity markets and/or placement of
both outward and/or inward conditions. In the context of the underwriting process itself,
the system allows for a fully automated customization of risk transfer pooling; i.e.,
insurance system solutions, offered in a controlled way, and which are simple, with
competitive wording and rates, a s well a s easy to administer and control. The system
allows, in a self-adapting way, to improve the "experience" in structuring and placing
risk transfer solutions; i.e., insurance and reinsurance systems. Finally, by means of the
constant monitoring and control of the pooled risk portfolio, the inventive system allows
for ensuring the operational capacity of the insurance system for covering losses
occurring a s a n impact related to transferred risk, thereby minimizing the required
pooled resources for the risk transfer. Therefore, the system also allows for monitoring the
effectiveness of automated (re)insurance systems, and for automatically optimizing
retention, required limits and provided coverage, etc. by providing self-adapting,
optimized and automated actuarial modeling, claims, accounting and contract
wording services.
In one embodied variant, the underwriting process flow of the underwriting
system is based on state-transition-controlled processing of workflow objects, and
wherein, by means of a control system, a n object is selected and processed following
the state-structured process flow comprising a plurality of process states, and wherein,
for each process state, one or more process tasks are executed by means of the
control system, and wherein the selected object is processed from one process state to
reach a subsequent process state. State parameters of a workflow object can, e.g., be
captured by capturing means of the control system, and a process state can, e.g., be
determined based on the captured state parameters, and they can be assigned to the
selected object, wherein, based o n the determined process state and/or state
parameters of the workflow object, at least one process task is generated by means of
the control system for a specific process state. Furthermore, a generated process task
can, e.g., be activated a s a function of task parameters that are assigned to a process
task. The process flow can, e.g., be dynamically operated by the control system,
wherein, by means of the control system, a n underwriting workflow object is processed
from the determined process state to reach a subsequent process state by executing
the assigned process tasks. The underwriting process flow can be, e.g., dynamically
generated and adapted, wherein the workflow object is processed by means of the
control system that initiates a subsequent process state by triggering defined trigger
values of the task parameters and/or state parameters of the preceding process state.
For the subsequent state transition within the process flow, upon measuring a no n
conforming workflow object, the subsequent process task can, e.g., be split into
subtasks, and wherein a subtask is generated by the control system to provide the
additional process states, which aretriggered and assigned to the process flow upon
arriving a t a measurement by the conformity index that exceeds the risk threshold
value, by means of the trigger module. This embodied variant has, inter alia, the
advantage that a n underwriting workflow object can be processed in a statestructured
underwriting process flow, wherein the state-structured process flow can be
fully controlled and operated by the monitoring and control system. Furthermore, it has
the advantage that applied process tasks of a process state of the underwriting
process flow can be further controlled by means of the assigned operating parameters,
wherein operational constraints or the splitting of specific tasks can be controlled by the
monitoring and control system by means of the operating parameters. The dynamic
assignment of operating parameters allows for a dynamic adaptation of the process
flow during the processing of a captured workflow object; i.e., a n object processed in
the underwriting process flow. This also has the advantage that the monitoring and
control system implemented a s a dynamically adaptable system can be automatically
optimized without any additional technical or human intervention. The present control
system for state-transition-controlled processing of underwriting workflow objects with
regard to responsive process management allows operators real-time visibility of their
processes (executing the same both within and externally to the platform), for modeling
and dynamically adapting their underwriting processes and associated transferred risks,
executing these processes by execution devices and appropriate signaling means to
those devices, sensing and responding to external events, and by incrementally
improving these processes. This is not possible with the systems a s known from the prior
art. Furthermore, this embodied variant has, inter alia, the advantage that any
processing of a n object can be handled fully automatically by means of the monitoring
and control system. In that way, the control system can automatically control, steer and
operate the underwriting processing of workflow objects within the process flow based
on the different state transitions of the selected object, and wherein the monitoring and
control system can, e.g., process the objects by means of steering and signal
transmission to the execution modules or devices.
In a further embodied variant, the control system comprises a n historical
engine device, wherein historical data of processed workflow objects associated with
known risk parameters are stored in a storing device of the control system, and wherein
the stored historical data are compared to the present workflow object, and whereby
relevant historical workflow object data are filtered from the stored data by means of a
filter-module, wherein the historic engine device and the control system are connected
by a data link for data signaling transmissions between the control system and the
historical engine device, and wherein the state-structured process flow is dynamically
generated by the control system, and the captured workflow object is dynamically
processed based on the data signaling transmission from the historical engine device
and the process management engine of the control system. This embodied variant has,
inter alia, the advantage that the control system provides a n improved process flow
based o n the comparison with historical data. This also allows for a n automatic
adaptation and optimization of the system and the generated process flow, which is
not possible in this manner by systems known from the prior art.
In another embodied variant, the underwriting process flow is a discrete
time-stochastical control process, wherein the control system comprises a stochastical
rating module, and wherein eachinitiation of the next process tasks is based a t least on
the selection of the process tasks of the preceding process state and a n additional
rating by means of the stochastical rating module. This embodiment variant has, inter
alia, the advantage that the monitoring and control system can generate and adapt
the process flow automatically and also steer and operate external devices by
appropriate signal generation, when a measurement is obtained indicating that the
conformity index exceeds the risk threshold value, and which is taken by means of the
trigger module.
In yet another embodied variant, the control system is self-adapting by the
automatic capturing of processed workflow objects after finishing the processing by
means of the underwriting process flow, wherein the content of the second database is
dynamically altered, storing the captured and processed workflow object by means of
the second searchable data structure of the second database. This embodied variant
has, inter alia, the advantage that any processing of a n object can be handled fully
automatically and without any further external interaction. Furthermore, the embodied
variant has the advantage that the monitoring and control system can be operated in
a self-adapting way, reacting automatically to the overall pooled risks by the
underwriting objects that are associated with the corresponding risk transfer.
In one embodied variant, control and steering signaling is generated by
means of a signaling module and transmitted to associated runtime execution modules
via the runtime interfaces, wherein the workflow object is processed by executing the
activated process tasks by means of the runtime execution modules based o n the
transmitted control and steering signaling, and wherein the process tasks are
generated by means of process task engines. This embodiment variant has, inter alia,
the advantage that non-conforming objects can be adjusted directly by means of the
monitoring and control system.
In one embodied variant, the automated underwriting system dynamically
adapts the overall risk that is pooled from the risk exposed components by adapting
the predefined boundary conditions that reflect the overall risk appetite of the
automated underwriting system. This embodied variant has, inter alia, the advantage
that the overall associated and pooled risk of a n automated insurance system can be
dynamically controlled and monitored by the automated insurance system by means
of the control and monitoring system.
In yet another embodied variant, the block elements comprise, in addition
to the one or more of interconnectable, search terms and/or meta-data, a variable
weighting parameter, and wherein each block elements and/or conformity thereof
with the stored block elements and with the recognized block elements with workflow
objects is weighted by the variable weighting parameter in order to generate the
conformity index. This embodied variant has, inter alia, the advantage that the
weighing parameters allow for incorporating functional relationships of the risk
associated with a specific underwriting object and/or a triggerable block element,
which is transferred to and pooled by the automated insurance system, and with the
incorporation of a specific block elements in the object.
According to the present invention, the above-mentioned objects for
controlling and monitoring of insurance underwriting systems for risk exposed
components are obtained, furthermore, in particular, in that a portfolio of risk transfer
objects of the underwriting system is monitored by consecutively capturing risk transfer
objects from the portfolio of pooled risk-transfer objects, and wherein a n activation of
additional process states is triggered upon the detection of non-conforming workflow
objects within the process flow by means of the control and monitoring system, and in
that a plurality of block elements is generated based on predefined boundary
conditions provided by a n automated underwriting system, and wherein the block
elements are triggerable parts of risk-transfer objects and comprise one or more of
interconnectable search terms and/or meta-data, and wherein the block elements are
assigned to a recognition block-map and stored by means of a first searchable data
structure provided by a first database, and in that the system comprises a second
database that provides a second searchable data structure for storing a plurality of risktransfer
objects, wherein the stored risk-transfer objects of the second database are a t
least partly generated based o n definable boundary condition parameters and/or
generated based on historical risk-transfer objects, and in that a risk-transfer object is
captured within the workflow pathway by means of the measurement and monitoring
system, and wherein the system comprises a core engine with a recognition module for
scanning the captured risk-transfer object thereby recognizing and identifying block
elements of the stored recognition block-map in the captured risk-transfer object, and
in that, based o n the filtered block elements of the captured risk-transfer object, a
proximity factor relative to each of the stored risk-transfer objects of the second
database is determined, and wherein a corresponding proximity factor is measured by
matching recognized block elements of the captured workflow object with block
elements of a workflow object of the second database, providing the measure for the
proximity of the two workflow objects based o n the mutually allocatable block
elements, and in that a conformity index is generated and assigned to the captured
risk-transfer object based on the conformity of the recognized block elements with the
stored block elements of the first searchable data structure and based o n the
conformity of the recognized block elements with risk-transfer objects of the second
database, and wherein only workflow objects within a predefined range of the
proximity factor are selected for generating the conformity index, and wherein the
conformity index provides a measure of the conformity of the captured workflow
object with the predefined boundary conditions of the automated underwriting system,
and in that the process states are triggered and assigned to the process flow by means
of a process management engine of the measurement and monitoring system, when
the conformity index is measured a s exceeding a risk threshold value, by means of the
trigger module. The captured workflow object can be content-scanned by means of
the recognition module for the purpose of recognizing and identifying block elements
by means of the recognition module , for example, and for triggering block elements of
the workflow objects by means of the stored block elements of the first searchable data
structure, and wherein triggered block elements of the captured workflow object are
classified by means of a classification module, and wherein associated entities are
extracted by means of a n entity extraction module, and wherein detected search
terms of the workflow object are flagged by means of a key-word flagging module. The
present embodiment of the invention has, inter alia, the same advantages a s the
preceding embodied variants of said invention. In particular, it allows for monitoring
and filtering the pooled risk-transfer objects; i.e., the pooled portfolio of risks by means
of the insurance system (20,21 ) in a fully automated manner, and, moreover, to
dynamically adapt the pooled risks by capturing objects based o n their rate of no n
conformity.
Finally, the present invention is not restricted to a system and method for a
pattern-recognition-based control and monitoring system for "insurance underwriting
systems" providing risk sharing for risk exposed components, but can explicitly be
applied a s a n appropriate system for the automated management, control and
monitoring of financial contracts. The invention generally can be realized for
automated systems for pattern-recognition-based control and monitoring to all kind of
data sets, comprising sensitive and recognizable data. Such systems and sensitive data
can e.g. comprise the control and monitoring of the following broad categories, i.e. (i)
financial data systems, wherein the types of financial data are numerous, but can e.g.
include credit card account numbers and tracking data, bank account numbers and
associated financial information, and a variety of credit-related data on individuals and
businesses. Such pattern-recognition-based control and monitoring financial systems
can e.g. also be provided for automated monitor and control the compliance or
consistency with various regulatory standards, a s e.g. Sarbanes-Oxley in the Unites
States, reporting financial data for public control systems. In that sense, the invention
also provides a pattern-recognition-based controlling and monitoring financial system
and/or a n automated financial compliance system; (ii) health data monitoring systems;
(iii) sensitive business data monitoring and controlling systems.
In addition to a system, a s described above, and a corresponding method,
the present invention also relates to a computer program product that includes
computer program code means for controlling one or more processors of the control
system in such a manner that the control system performs the proposed method; and it
relates, in particular, to a computer program product that includes a computerreadable
medium containing therein the computer program code means for the
processors.
Brief Description of the Drawings
The present invention will be explained in more detail below by way of
example, with reference to the drawings in which:
Figure 1 shows a block diagram illustrating schematically a n exemplary
system according to the invention for a pattern-recognition based control and
monitoring system 10 for insurance underwriting systems 20, 2 1 for risk exposed
components 5 1, 52, 53, 54, wherein a n underwriting workflow of the underwriting system
20, 2 1 and/or the pooled risk-transfer objects are monitored by capturing underwriting
objects 7 1, 72, 73 in the underwriting workflow 19 and/or by selecting one of the pooled
underwriting objects 7 1, 72, 73 from the risk-transfer systems 20,21 . An activation of
additional process states 19 11, 1 912,1 913 is triggered upon the detection of no n
conforming workflow risk-transfer objects 7 1, 72, 73 within the process flow 19 and/or
risk-transfer objects 7 1, 72, 73 selected from the pooled risk-transfer objects by means of
the control and monitoring system 10.
Figure 2 shows a block diagram illustrating schematically a n exemplary
technical back-end framework of the recognition modulel 32 providing autoclassification,
entity extraction and flagging of documents that match one or more
scenarios, a s defined by the control and monitoring system 10 by means of the
classification module 133, entity extraction module 134, and key-word flagging module
135. The structure of the recognition modulel 32 is modular and flexible; and it is to be
easily adaptable to the requirements involved in changing underwriting and risktransfer
condition parameters of different automated risk-pooling systems 20,21 and risk
exposed components 50,51 ,52. Therefore, the structure is independent of specific
underwriting conditions. The recognition module 132 is implemented, in particular,
based on machine learning (ML) and/or statistical learning and/or bioengineering. The
monitoring and control system 10 enables process automation through suggestive
classification, auto-filing and notification of underwriting objects 7 1,72,73 related to one
or more scenarios; i.e., based on predefined risk-transfer and/or underwriting boundary
or condition parameters.
Figure 3 shows a block diagram illustrating schematically a n exemplary
recognition flow a s implemented by the recognition modulel 32. The modular
framework allows for plug-in and plug-out of new data processing se†upsand/or
algorithms and different applications with low implementation efforts. To improve
flexibility, the signaling by means of the signaling module 1 can, e.g., be provided in a
standard format using XML. The recognition modulel 32 can provide the block of text of
target objects 7 1,72,73 with a very high degree of matching.
Figure 4 shows a block diagram illustrating schematically a n exemplary
classification process, a s implemented by means of the classification module 133.
Figure 5 shows a block diagram illustrating schematically a n exemplary
entity extraction process, a s implemented by means of the entity extraction module
134.
Figure 6 shows a block diagram illustrating schematically an exemplary key
word flagging process, a s implemented by means of the key-word flagging module
135.
Figure 7 shows a block diagram illustrating schematically a n exemplary
string alignment process, a s implemented by means of the recognition module 132.
Figure 8 shows a block diagram illustrating schematically a n exemplary
global alignment process and a local alignment process, a s implemented by means of
the recognition module 132.
Figure 9 shows a block diagram illustrating schematically a n exemplary
comparison process based on a dot-plot matrix, as, e.g., implemented by means of the
recognition module 132.
Figure 10 shows a block diagram illustrating schematically a n exemplary
part of the recognition process, a s implemented by means of the recognition module
132.
Detailed Description of the Preferred Embodiments
Figure 1 schematically illustrates a n architecture for a possible
implementation of an embodiment of the pattern-recognition based control and
monitoring system 10 for insurance underwriting systems 20, 2 1 for risk exposed
components 5 1, 52, 53, 54 and/or for automated portfolio management systems of risktransfer
underwriting objects. In Figure 1, reference numeral 10 refers to the inventive
monitoring and control system. The monitoring and control system 10 is implemented
based on underlying electronic components, steering codes and interactive interface
devices such as, e.g., signal generation modules, or other module that interact
electronically by means of the appropriate signal generation between the different
modules, devices, or the like. An object 7 1,72,73 that is captured for controlling and
monitoring by means of system 10 can comprise, e.g., a t least one workflow object
7 1,72,73, which is processed in a n automated underwriting process flow 19 enabling a
specifically defined risk-transfer between the risk exposed components 5 1, 52, 53, 54
and the automated resource pooling system 20,21 , i.e. the automated insurance
system, wherein the object 7 1,72,73 is captured during the processing of the object
7 1,72,73 within the pathway of the underwriting workflow 19 or after finishing the
automated workflow, i.e. processing of the underwriting object 7 1,72,73, by capturing
the object(s) 7 1,72,73 from a source of already processed and pooled risk-transfer
7 1,72,73 of a n automated underwriting and/or insurance system 20,21 . An object
7 1,72,73 that is captured for controlling and monitoring by means of system 10 can also
merely comprise, e.g., a risk-transfer object 7 1,72,73 from a source of pooled risk-transfer
objects 7 1,72,73 of the automated insurance system 20,21 ; i.e., from the portfolio of risktransfer
objects 7 1,72,73 a s pooled by the resourse pooling systems 20,21 . In these
cases, the monitoring and control system 10 is able to provide monitoring and filtering
of non-conforming risk-transfer objects or underwriting objects 7 1,72,73, and activating
of the appropriate signaling and/or further processing and adjusting and/or separating
of these non-conforming measured objects 7 1,72,73. An object 7 1,72,73 is selected by
means of a selecting or filtering module using the monitoring and control system 10;
and it is processed following a recognition process a s illustrated in Figures 2 to 10.
When the object 7 1,72,73 is captured in the underwriting workflow 19 of the
underwriting system 20, 2 1, i.e. the object 7 1,72,73 is a processed workflow object
7 1,72,73 of a n underwriting process of the automated insurance systems 20,21 , the
workflow 19 is monitored and controlled by capturing workflow objects 7 1, 72, 73 and
by triggering activation of additional process states 191 1,1912,1913 upon the detection
of non-conforming workflow objects 7 1, 72, 73 within the process flow 19 by means of
the control and monitoring system 10. However, the monitoring and control system 10
can also be applied to already processed and pooled objects 7 1,72,73, particularly by
means of the automated insurance system 20,21 , and wherein the overall risk that is
pooled by the automated insurance system 20,21 is given by the overall pooled risktransfer
object 7 1,72,73; e.g., stored and assigned to the insurance systems 20,21 in a n
appropriate storage source of the automated insurance systems 20,21 . In this case, the
monitoring and control system 10 monitors and controls the corresponding portfolio of
risk transfer objects 7 1,72,73 of the underwriting system 20, 2 1 by consecutively
capturing risk transfer objects 7 1, 72, 73 of the portfolio of pooled risk-transfer objects 7 1,
72, 73 from the storage source, and wherein a n activation of process states
191 1,1912,1913 is triggered upon the detection of non-conforming risk-transfer objects
7 1, 72, 73 within the portfolio of pooled risk-transfer objects 7 1, 72, 73 by means of the
control and monitoring system 10. In the following, the references to the objects
7 1,72,73 a s workflow objects 7 1,72,73 or pooled risk-transfer objects 7 1,72,73 are used
interchangeably, and wherein a person skilled in the art understands that the control
and monitoring system 10 can be applied in the above-described ways by way of
monitoring objects 7 1,72,73 either during the underwriting process, or directly after the
underwriting process, or a s already processed objects 7 1,72,73 pooled in a portfolio of
a source assigned to the automated insurance system 20,21 .
For the implementation of the controlled underwriting process, for each
underwriting process state 191 1, 1912, 1913, one or more underwriting process tasks
1921 , 1922, 1923 can be executed by the monitoring and control system 10 in order to
process the workflow object 7 1,72,73 from one process state 191 1, 191 2, 19 13 to a
subsequent process state 191 1, 191 2, 1913. Changes in the process flow can be
induced by the execution of one or a plurality of tasks. However, process and tasks can
be independently implemented in the system, wherein possible relations between
processes or process flows and the execution of one or more tasks are implemented a s
constraints. The process tasks 1921 , 1922, 1923 do not have to be necessarily generated
by the control and monitoring system 10; but they can, a s in the embodied variant, also
be inserted or imported from external sources, such a s denoted databases with
appropriate, predefined process tasks or input means, such a s consoles for manual
inputs, etc.
The underwriting process or workflow 19 comprises the technical and/or
procedural steps required for executing the controlled processing of objects 7 1,72,73, in
particular a state transition-controlled processing by means of the control and
monitoring system 10. Furthermore, the system 10 comprises technical and other means
for conducting the processing steps, and the transfer and flow of data/signaling
between the means and/or steps for executing the process with regard to the objects
7 1,72,73. The objects 7 1,72,73 are processed by a set of processes or tasks that need to
be carried out. Within a n underwriting process flow 19, underwriting objects 7 1,72,73
pass through the different tasks 1921 , 1922, 1923 and process states 191 1, 1912, 1913 in
the specified order, from start to finish; and the tasks 1921 , 1922, 1923 are executed
either by dedicated technical processing devices or means, or by specified functions
that are generated by the monitoring and control system 10 instructing a processor
device, or by dedicated signaling generation for the purpose of instructing performing
activities/tasks o n the underwriting workflow object 7 1,72,73. The monitoring and control
system 10 comprises a process management engine 13 for executing the underwriting
process flow 19. The generation of a specific process flow 19, i.e. the process
management process definition, can be generated, e.g., based on a desired process
in a process management generator of the process management engine 13. As
described below, the process flow 19 can be generated dynamically (state by state)
based upon a t least measuring parameters of the capturing means 15,151 154
and/or data transmitted via the interface module 18 and the input devices 181 ,182,183,
e.g., via input by the insurance systems 20,21 or the risk exposed components 5 1,52,53.
In one embodied variant, the generated process flow 19 can, e.g., be dynamically or
partly dynamically translated into a processor source code, such as, e.g. Java source
code or the like, at a translator engine of the process management engine 13. The
source code can then be compiled into a byte code in the context of a compiler
engine of the process management engine 13. Finally, a virtual machine of the
processing device or a processor-driven, -steered or -operated device of the
monitoring and control system 10 can be configured for executing the byte code. Such
devices can comprise execution devices of the process tasks 1921 , 1922, 1923, such as,
e.g., the runtime execution modules 30,31 . Therefore, the process flow 19 is modeled
and generated by means of the process management engine 13, including or based
upon specific processing rules and technical instructions stored in a database. A person
skilled in the art understands that the scope of the technical instructions must has to be
constructed along broad lines, comprising all technically necessary information, data,
specifications or operational parameters, such a s to allow the processing rules to be
executed by the system 10. Furthermore, at least parts of the parameters of the
underwriting objects 7 1, 72, 73 can, e.g., be captured via a plurality of input devices
181 ,182,183 that are accessible to risk exposed components 5 1,52,53, and/or
automated insurance systems 20,21 for dynamically adapting the workflow objects
7 1,72,73 via the interface module 18 and/or network 60, based o n the appropriate
signal generation by means of the signaling module 1 of the control system 10. The
network 60 can include a hard-wired or wireless network; e.g., the internet, a GSM
network (Global System for Mobile Communication), a n UMTS network (Universal Mobile
Telecommunications System) and/or a WLAN (Wireless Local Region Network), and/or
dedicated point-to-point communication lines.
The underwriting processing of the object 7 1,72,73 or the portfolio of pooled
objects 7 1,72,73 can, e.g., be monitored by means of dedicated monitoring and/or
measuring devices 15,151 ,152,1 53,154 of the monitoring and control system 10,
particularly based o n at least captured state parameters 191 1, 1912, 1913 or the taskparameters
1921 , 1922, 1923 of the underwriting workflow 19 of the object 7 1,72,73
and/or risk-transfer parameters that are associated with the underwriting object
7 1,72,73,74. The monitoring and control system 10 can be implemented using one or a
plurality of visual front ends. The execution of the additional process states 191 1, 1912,
19 13 or the process tasks 1921 , 1922, 1923 is controlled and/or steered and operated by
means of the monitoring and control system 10 or by a dedicated process flow 19
execution engine that handles the call-up and signal generation of the remote devices
or applications. For the additional underwriting process states 191 1, 1912, 1913,
executed non-conforming captured objects 7 1,72,73, the monitoring and control
system 10 can, e.g., further comprise a double-tier structure, wherein the first-tier
execution of the control system 10 comprises the generation of the additional process
states 191 1, 1912, 19 13 and the second-tier execution of the control system 10
comprises the generation and association of the process tasks 1921 , 1922, 1923, wherein
the processing of the object 7 1,72,73 and the appropriate signaling 141 ,142,143 by the
monitoring and control system 10 is dynamically adapted based on at least the
alterable operating parameters of the associated process states 19 11, 1912, 19 13 or
tasks 1921 , 1922, 1923. The reference numeral 141 represents the appropriate signaling
dedicated to the runtime execution module 30, 142, the signaling for runtime execution
module 3 1, and so forth. Underwriting process state parameters of the captured object
7 1,72,73 can be captured by capturing means 151 ,152,153,1 54 of the monitoring and
control system 10, and a process state 191 1, 1912, 1913 is determined based on the
captured state parameters. The determined process state 191 1, 1912, 1913 is assigned
to the selected object 7 1,72,73 by the monitoring and control system 10. Upon
triggering non-conformity of a n object 7 1,72,73, based on the determined process state
191 1, 1912, 1913 and/or state parameters of the captured object 7 1,72,73, a t least one
process task 1921 , 1922, 1923 is generated by means of the monitoring and control
system 10, wherein, for a specific process state 191 1, 191 2, 1913, a generated process
task 1921 , 1922, 1923 is activated a s a function of the task-parameters assigned to a
process task 1921 , 1922, 1923. The control system 10, e.g., can generate one or more
process tasks based on the process state 191 1, 1912, 19 13 of the selected object
7 1,72,73 in order to bring the object 7 1,72,73 in a predefinable rate of conformity.
Measuring operations for the rate of conformity are described in detail below.
The monitoring and control system 10 generates a plurality of block
elements 1112, 1113, 1114 based on predefined boundary conditions provided by a n
automated underwriting system 20, 2 1. The block elements 1112, 1113, 1114 form
triggerable parts of workflow and/or underwriting objects 7 1, 72, 73 and comprise one
or more of interconnectable search terms and/or meta-data, and wherein the block
elements 1112, 1113, 1114 are assigned to a recognition block-map 111 and stored by
means of a first searchable data structure 112 provided by a first database 11. The
plurality of block elements 1112, 1113, 1114 can, e.g., be generated, inter alia, based
on predefined boundary conditions provided by an automated underwriting system 20,
2 1 and by at least partially extracting data from historical and/or simulated data.
Furthermore, the one or more of the interconnectable search terms and/or meta-data
can, e.g., be extracted and/or generated at least partially from historical and/or
simulated data. In addition, the block elements 1112, 1113, 1114 can, e.g., comprise the
one or more of the interconnectable search terms and/or meta-data, a variable
weighting parameter, and wherein each block element 1112, 1113, 1114 and/or its
conformity with the stored block elements 1111, 1112, 1113 and with the recognized
block elements with workflow objects 121 1, 12 12 is weighted by the variable weighting
parameter in order to generate the conformity index. The weighing parameter allows
for incorporating functional relationships of the risk that is associated with a specific
underwriting object 7 1,72,73 and/or a triggerable block element 1111, 1112, 1113,
which is then transferred to and pooled by the automated insurance system 20,21 with
the incorporation of specific block elements 1111, 1112, 1113 in the object 7 1,72,73.
The system 10 comprises a second database 12 that provides a second
searchable data structure 12 1 for storing a plurality of workflow objects 121 1, 1212,
wherein the stored workflow objects 121 1, 12 12 of the second database 12 are at least
in part generated based on definable boundary condition parameters and/or historical
workflow objects. The stored workflow objects 121 1, 12 12 of the second database 12
can, e.g., be at least in part generated based on definable boundary condition
parameters selecting block elements 1112, 1113, 1114 from the first database 11 and/or
extracting workflow objects from historical workflow objects' data.
A workflow object 7 1, 72, 73 is captured within the workflow pathway 19 or
from a portfolio of risk-transfer objects 7 1,72,73, which is stored in a storage source of
the insurance system 20,21 , by means of the measurement and monitoring system 10.
For the captured underwriting objects 7 1,72,73, the system 10 comprises a core engine
131 with a recognition module 132 for scanning the captured workflow objects 7 1, 72,
73, thereby recognizing and identifying block elements 1112, 1113, 1114 of the stored
recognition block-map 111 in the captured workflow objects 7 1, 72, 73. The captured
workflow object can, for example, be content-scanned by means of the recognition
module 132 for recognizing and identifying block elements 112, 113, 114 by means of
the recognition module 132, thereby triggering block elements of the workflow objects
7 1, 72, 73 by means of the stored block elements of the first searchable data structure
111. Triggered block elements of the captured workflow object 7 1,72,73 are classified
by means of a classification module 133. Further associated entities are extracted by
means of an entity extraction module 134 and detected search terms of the workflow
object are flagged by means of a key-word flagging module 135.
Based on the filtered block elements of the captured workflow object 7 1,
72, 73, a proximity factor regarding each of the stored workflow objects 121 1, 12 12 of
the second database 12 is determined, wherein a corresponding proximity factor is
measured by matching recognized block elements of the captured workflow object 7 1,
72, 73 with block elements 12 11, 12 12 of a workflow object 12 11, 12 12 of the second
database 12, thereby providing the measure for the proximity of the two workflow
objects based on the mutually allocatable block elements. The corresponding proximity
factor can, e.g., be measured by matching recognized block elements of the
captured workflow object 7 1, 72, 73 with block elements 121 1, 12 12 of a workflow
object of the second database 12, thus providing the measure for the proximity of the
two workflow objects based on the mutually allocatable block elements and/or their
degree of conformity. Furthermore, based on the filtered block elements 12 11, 12 12 of
the captured workflow object 7 1, 72, 73, the proximity factor regarding each of the
stored workflow objects 121 1, 1212 of the second database 12 can, e.g., be
determined, and wherein a corresponding proximity factor is measured by matching
recognized block elements of the captured workflow object 7 1, 72, 73 with block
elements of a workflow object 121 1, 12 12 of the second database 12, thus providing
the measure for the proximity of the two workflow objects based o n the number of
mutually allocatable block elements and/or their degree of conformity.
A conformity index is generated and assigned to the captured workflow
object 7 1, 72, 73, respectively based o n the conformity of the recognized block
elements with the stored block elements 1111, 1112, 1113 of the first searchable data
structure 111 and based o n the conformity of the recognized block elements with
workflow objects 121 1, 12 12 of the second database 12, and wherein only workflow
objects 121 1, 1212 within a predefined range of the proximity factor are selected for
generating the conformity index. The conformity index provides a measure for the
conformity of the captured workflow object 7 1, 72, 73 within the predefined boundary
conditions of the automated underwriting system 10. The conformity index can, for
example, be generated and assigned to the captured workflow object 7 1, 72, 73
based o n the conformity of the recognized block elements with the stored block
elements 1111, 1112, 1113 of the first searchable data structure 111 and based on the
conformity of the recognized block elements with workflow objects 121 1, 12 12 of the
second database 12, wherein the conformity index provides a measure for the
variance of the captured workflow object 7 1, 72, 73 within the predefined boundary
conditions of the automated underwriting system 10.
The additional process states are triggered and assigned to the process flow
19 by means of a process management engine 13 of the measurement and monitoring
system 10, when the conformity index is measured a s exceeding a risk threshold value
141 a s established by means of the trigger module 14. To provide the required signaling
and self-driven interaction of the control and monitoring system 10 with other devices,
control and steering signaling 141 ,142,143 is generated by means of a signaling module
14 and, for example transmitted to associated runtime execution modules 30,31 via the
runtime interfaces 3 11, 3 12, wherein the workflow object 7 1,72,73 is processed by
executing the activated process tasks 1921 , 1922, 1923 by means of the runtime
execution modules 30,31 based o n the transmitted control and steering signaling
14 1,142, 143, and wherein the process tasks 1921 , 1922, 1923 are generated by means of
process task engines 312, 322. Based on the signaling of the monitoring and control
system 10, the automated underwriting system 20, 2 1 can, e.g., be dynamically
adapting, in particular in a self-adapting way; a n overall risk that is pooled from the risk
exposed components 5 1 54 by adapting the predefined boundary conditions
reflecting the overall risk appetite of the automated underwriting system 20, 2 1. This
level of self-adjustment and automated control is not reached by any of the known
systems of the prior art.
In one embodied variant, the underwriting process flow 19 of the
underwriting system 20, 2 1 is based o n state-transition-controlled processing of workflow
objects 7 1,72,73. In this embodied variant, a n object 7 1,72,73 is selected by means of
the monitoring and control system 10 and processed following the state-structured
process flow 19 comprising a plurality of process states 191 1,1912,1913; and for each
process state 191 1,1912,1 913 one or more process tasks 1921 ,1922,1923 are executed
by means of the control system 10, and wherein the selected object 7 1,72,73 is
processed from one process state 191 1, 1912,191 3 to a subsequent process state
191 1,1912,1913. State parameters of a workflow object 7 1,72,73 can, e.g., be captured
by capturing means 15,151 ,152,153,154 of the monitoring and control system 10; and a
process state 191 1,1912,1913 is determined based on the captured state parameters
and assigned to the selected object 7 1,72,73. Based on the determined process state
191 1,1912,1913 and/or state parameters of the workflow object 7 1,72,73, at least one
process task 1921 , 1922, 1923 can be generated by means of the monitoring and
control system 10 for a specific process state 191 1,1912,1913. The generated process
task 191 1, 1921 , 1923 can, e.g., be activated a s a function of the task parameters
assigned to a process task 1921 , 1922, 1923. The capturing means 15 of the monitoring
and control system 10 can comprise measuring devices and/or measuring sensors
151 ,152,153,154 for capturing the state parameters and/or task parameters of the
workflow object 7 1,72,73.
In another embodied variant, the process flow 19 is dynamically operated
by the monitoring and control system 10; wherein by means of the monitoring and
control system 10, a n underwriting workflow object 7 1,72,73 is processed from the
determined process state 121 ,122,123 to a subsequent process state 191 1,1912,1913 by
executing the assigned process tasks 1921 , 1922, 1923, 1924. The underwriting process
flow 19 can, e.g., be dynamically generated and adapted, wherein the workflow
object 7 1,72,73 is processed by means of the control system 10 that initiates a
subsequent process state 191 1,1922,1933 by triggering defined trigger values of the task
parameters and/or state parameters of the preceding process state 191 1,1922,1933. For
the subsequent state transition within the process flow 19, the subsequent process task
191 1, 1912, 1923 can, e.g., be split by means of subtasks, wherein a subtask is
generated by the monitoring and control system 10 to provide the additional process
states triggered and assigned to the process flow 19, when arriving a t a measurement
of the conformity index that exceeds the risk threshold value 141 by means of the
trigger module 14. Finally, a s a n additional embodied variant, the monitoring and
control system 10 comprises a historical engine device 1 , wherein historical data of
processed workflow objects associated with known risk parameters are stored in a
storing device of the control system 10, and wherein the stored historical data are
compared to the present workflow object 7 1, 72, 73, and relevant historical workflow
object 7 1, 72, 73 data are filtered from the stored data by means of a filter-module 17,
wherein the historical engine device 1 and the monitoring and control system 10 are
connected by a data link for data signaling transmission between the control system 10
and the historical engine device 1 , and wherein the state-structured process flow 19 is
dynamically generated by the control system 10 and the captured workflow object is
dynamically processed based o n the data signaling transmission from the historical
engine device 1 and the process management engine 13 of the control system 10. As
mentioned, historical data of past underwriting objects 7 1,72,73 and of underwriting
processes 19 are stored in a storing device of the monitoring and control system 10. The
stored historical data are compared to the captured underwriting object 7 1,72,73;
then, relevant historical object data and/or process flow data from the stored data are
filtered by means of a filter module 17.
When a signal of non-conformity has been triggered within a predefinable
range by means of the control and monitoring system 10, the monitoring and control
system 10 and the runtime execution devices 30,31 can, e.g., interact in runtime,
wherein the object 7 1,72,73 is processed based on the dynamically adapted process
flow 19 with the generated process tasks 1921 , 1922, 1923 by executing the activated
process tasks 1921 , 1922, 1923 by means of the runtime execution devices 30,31 based
on the transmitted control and steering signaling 141 ,142,143. For example, a first-tier
execution of the control system 10 can comprise the generation of the process states
191 1,1912,1913 and a second-tier execution of the control system 10 can comprise the
generation and association of the additional process tasks 191 1,1912,1913, when a
measurement has been taken, whereby the conformity index exceeds the risk threshold
value 141 , by means of the trigger module 14, and wherein the processing of the
workflow object 7 1,72,73 and the appropriate signaling 141 ,142,143 by the control
system 10 is dynamically adjusted by alternating operating parameters of the runtime
execution devices 30,31 . Furthermore, the monitoring and control system 10 can, e.g.,
be self-adapted by automatically capturing of processed workflow objects 7 1, 72, 73
after finishing the processing by means of the underwriting process flow 19, and wherein
the content of the second database 12 is dynamically altered, storing the captured
and processed workflow object 121 1, 12 12 by means of the second searchable data
structure 121 of the second database 12.
At least parts of the task parameters of the process tasks 1921 , 1922, 1923 of
the additional process states 191 1, 1912, 1913 can be dynamically generated by the
control system 10 based on the measured conformity index of the captured workflow
object 7 1,72,73. An appropriate signaling 141 ,142,143 can, e.g., be generated by
means of the signaling module 14 for steering the execution devices 50,51 ,52
processing the selected object 7 1,72,73 according to the process flow and/or
additional process flow 19 generated by the control system 10. Finally, the processing of
the captured workflow object 7 1,72,73 can, e.g., be monitored by means of dedicated
monitoring and/or measuring devices 30, 3 1 of the control system 10, particularly based
on at least the captured state parameters, the tasks parameters and/or operating
parameters.
As an embodiment variant, the workflow object 7 1, 72, 73 automatically
can be assembled within the workflow pathway 19 by means of the additional process
states, which are triggered and assigned to the process flow 19 by means of the
process management engine 13 of the measurement and monitoring system (10),
wherein by means of the system 10 based on a predefined structure by means of
composing a plurality of block elements 1112, 1113, 1114 triggered within the first
database 11.
In a n even further embodiment variant at least one risk-importing
recognized block element of the workflow object 7 1, 72, 73 is substituted by block
elements 1112, 1113, 1114 triggered within the first database 11 improving the
measured conformity index, wherein the block element of the workflow object 7 1, 72,
73 are substituted by means of the additional process states, which are triggered and
assigned to the process flow 19 by means of the process management engine 13 of the
measurement and monitoring system 10. Risk-importing recognized block elements of
the workflow objecf 7 1, 72, 73 can automatically be substituted by block elements
1112, 1113, 1114 triggered within fhe first database 11 by means of fhe sysfem 10, until
fhe measured conformity index of fhe captured workflow objecf 7 1, 72, 73 is measured
within fhe predefined boundary conditions of fhe automated underwriting sysfem 10.
In summary, the sysfem 10 provides a n automated device for a broad
range of automated functionalities, a s e.g. (i) automated document review
functionalities, a s e.g. automatical display of critical comments when a stored block
element, a s a clause, is defected in a workflow objecf 7 1, 72, 73 (e.g. a document). The
sysfem 10 can e.g. generate visual metaphors for creating a "risk map" of a workflow
object 7 1, 72, 73. The sysfem 10 can also transmit the workflow objecf 7 1, 72, 73 to
contract reviewers adapting fhe automatic result, and transform it info a generated
contracts review; (ii) automated treasury functionalities, a s the sysfem 10 comprises the
database 11 a s a storage point providing and containing a n up-to-date collection of
all clauses in a portfolio, together with user-specific data, e.g. remarks, comments or
views, o n each. This feature can also be used a s searchable platform for training a s well
a s for ensuring consistency; (iii) automated check listing functionalities, a s e.g.
automatic generation of checklists of content, e.g. exclusion. This also can comprise
automated generation of variable checklists per LoB (Line of Business) and/or per
scope, a s required e.g. by different underwriter groups, which has fhe advantage of
saving work time of users by means of fhe automated generation; (iv) automated
document comparison functionalities, a s e.g. a workflow objecf 7 1, 72, 73, for example
contract data, is captured a n compared across a user specific portfolio or stored
collection. In this way, similar workflow objecf 7 1, 72, 73 of fhe second database 12 are
defected by fhe system 10 and fhe captured workflow objecf 7 1, 72, 73 automatically
can be compared, in particular providing a measurable proximity factor and
conformity index; (v) automated document search functionalities, a s e.g. a free text
search engine can be integrated into fhe system 10 providing free search for a
workflow objecf 7 1, 72, 73 or a block element. For example, upon entering e.g.
"Endorsement 5 WXL Liability Client X 201 3 draff 3" in the integrated search engine, the
sysfem will create a link †o fhe corresponding workflow objecf 7 1, 72, 73 or block
element; (vi) automated content search functionalities, a s e.g. a free text search
engine can be integrated into fhe system 10 providing free search for block elemenf a s
stored in fhe first and/or second database 11/12. Allowing content searches across the
whole stored, workflow objects 7 1, 72, 73 or block elements, e.g. specified segments or
inducted documents within a portfolio overview of a specific user; (vii) automated
document benchmarking functionalities providing standardization indexing and
automated workflow objects 7 1, 72, 73, e.g. contract, summaries by means of the
system 10. This feature also allows benchmarking and segmentation for standard and
non standard workflow objects 7 1, 72, 73; (viii) automated document drafting
functionalities, wherein the system 10 assembles a s workflow object 7 1, 72, 73 e.g. a
draft contract based on a structured predefined "order form" while selecting only
approved clauses for block elements; (ix) automated document re-drafting
functionalities automatically improving captured workflow objects 7 1, 72, 73, for
example inputting documents a s workflow objects 7 1, 72, 73 assembled by third parties,
thereby automatically substituting detected risk-importing block elements, a s e.g. riskimporting
clauses, by approved block elements. For control reasons, the substituted
workflow objects 7 1, 72, 73 can e.g. be marked or otherwise labeled by the system 10.
The automated insurance systems 20,21 are implemented a t least
comprising a n automated resource pooling system for pooling resources of the risk
exposed components 5 1,52, 53, 54 in order to provide risk protection for the risk
exposed components 5 1 54 by means of the pooled resources based on the pooled
risks associated with a underwriting object 7 1,72,73. The pooled resources can be, e.g.,
based o n monetary parameters. The resource-pooling systems 20,21 can, e.g.,
comprise all the necessary technical means for electronic money transfers and
association, as, e.g., initiated by one or more associated payment transfer modules via
a n electronic network 60. The monetary parameters can be based o n any available
electronic and transferable means, such as, e.g., e-currency, e-money, electronic cash,
electronic currency, digital money, digital cash, digital currency, or cyber currency,
etc., which can only be exchanged electronically. A payment data store of the
automated insurance systems 20,21 can, e.g., provide the means for associating and
storing monetary parameters, which are linked to a single of the pooled risk exposed
components 5 1,52,53,54. The present invention can involve the use of the mentioned
network 60, such as, e.g., computer networks or telecommunications networks, and/or
the internet and digitally stored value systems. Electronic funds transfer (EFT), direct
deposit, digital gold currency and virtual currency are further examples of electronic
money. Also, the transfer can involve technologies, such a s financial cryptography and
technologies enabling the same. For the transaction of the monetary parameters, it is
preferable that hard electronic currency be used, without the technical possibilities for
disputing or reversing any charges. The resource-pooling system 20,21 can support, for
example, non-reversible transactions. The advantage of this arrangement is that the
operating costs of the electronic currency system are greatly reduced by not having to
resolve payment disputes. However, this way, it is also possible for electronic currency
transactions to clear instantly, making the funds available immediately to the
automated insurance systems 20,21 . This means that using hard electronic currency is
rather akin to a cash transaction. However, also conceivable is the use of soft
electronic currency, such a s currency that allows for the reversal of payments, for
example, having a "clearing time" of 72 hours, or the like. The modality of the electronic
monetary parameter exchange applies to all connected systems and modules related
to the resource-pooling systems 20,21 of the present invention, such as, e.g.,
appropriate payment transfer modules.
List of reference signs
10 Control and monitoring system
11 First database
111 Recognition block-map
1111, 1112, 1113 Triggerable block elements
112 First searchable data structure
12 Second database
121 Second searchable data structure
121 1, 1212 Stored workflow objects
13 Process management engine
131 Core engine
132 Recognition module
133 Classification module
134 Entity extraction module
135 Key-word flagging module
136 Trigger module
1361 Threshold value
14 Signaling module
141 ,142,143 Control and steering signaling
15 Capturing means
5 , 52, 153, 54 Measuring devices and/or sensors
1 Historical engine device
17 Filter module
18 Interface module
181 ,182,183 Input device
19 Process flow (state-structured)
191 1,1912,1913 Process state
1921 ,1922,1923 Process task
20, 2 1 Insurance underwriting systems
3 1, 32 Runtime execution devices
3 11, 321 Signaling interface
321 , 322 Process task engine
50,51 ,52 Risk exposed components
Network
,72,73 Underwriting objects of the process flow or an object-source
Claims
1. A pattern-recognition based method for a control and monitoring system
(10) for automated underwriting systems (20, 2 1) or automated data caputing and
surveillance systems, wherein the workflow of a n underwriting system (20, 2 1) and/or a n
object-source of a n underwriting system (20, 2 1) is monitored by capturing underwriting
objects (71 , 72, 73) in the underwriting process flow (19) or from the object-source, and
wherein a n activation of additional process states (191 1,1912,191 3) is triggered upon
the detection of non-conforming workflow objects (71 , 72, 73) by means of the control
and monitoring system (10), characterized
in that a plurality of block elements ( 1 112, 1113, 1114) are generated based
o n predefined boundary conditions provided by a n automated underwriting system
(20, 2 1) , and wherein the block elements ( 1 112, 1113, 1114) are triggerable parts of
workflow objects (71 , 72, 73) and comprise one or more of interconnectable search
terms and/or meta-data, and wherein the block elements ( 1 112, 1113, 1114) are
assigned to a recognition block-map ( 1 11) and stored by means of a first searchable
data structure ( 1 12) provided by a first database ( 1 1) ,
in that the system (10) comprises a second database (12) providing a
second searchable data structure (121 ) for storing a plurality of workflow objects (121 1,
1212), and wherein the stored workflow objects (121 1, 1212) of the second database
(12) are a t least partly generated based o n definable boundary condition parameters
and/or historical workflow objects,
in that a workflow object (71 , 72, 73) is captured within the workflow
pathway (19) by means of the measurement and monitoring system (10), and wherein
the system (10) comprises a core engine (131 ) with a recognition module (132) for
scanning the captured workflow object (71 , 72, 73), thereby recognizing and identifying
block elements ( 1 112, 1113, 1114) of the stored recognition block-map ( 1 11) in the
captured workflow object (71 , 72, 73),
in that, based o n the filtered block elements of the captured workflow
object (71 , 72, 73), a proximity factor relative to each of the stored workflow objects
(121 1, 1212) of the second database (12) is determined, and wherein a corresponding
proximity factor is measured by matching recognized block elements of the captured
workflow object (71 , 72, 73) with block elements of a workflow object (121 1, 1212) of the
second database (12), providing the measure for the proximity of the two workflow
objects based o n the mutually allocatable block elements,
in that a conformity index is generated and assigned to the captured
workflow object (71 , 72, 73) based o n the conformity of the recognized block elements
with the stored block elements ( 1 111, 1112, 1113) of the first searchable data
structure ( 1 11) and based on the conformity of the recognized block elements with
workflow objects (121 1, 1212) of the second database (12), and wherein only workflow
objects (121 1, 121 2) within a predefined range of the proximity factor are selected for
generating the conformity index, and wherein the conformity index provides a measure
for the conformity of the captured workflow object (71 , 72, 73) with the predefined
boundary conditions of the automated underwriting system (10), and
in that the additional process states are triggered and assigned to the
process flow ( 19) by means of a process management engine (13) of the measurement
and monitoring system (10), when the measurement of the conformity index exceeds a
risk threshold value (141 ) , a s establiahed by means of the trigger module (14).
2. The method according to claim 1, characterized in that, for recognizing
and identifying block elements ( 1 12, 113, 114) by means of the recognition module
(132), the captured workflow object is content-scanned by means of the recognition
module ( 1 32) triggering for block elements of the workflow objects (71 , 72, 73) by
means of the stored block elements of the first searchable data structure ( 1 11) , wherein
triggered block elements of the captured workflow object are classified by means of a
classification module (133), wherein associated entities are extracted by means of a n
entity extraction module ( 134), and wherein detected search terms of the workflow
object are flagged by means of a key-word flagging module (135).
3. The method according to one of the claims 1 to 2, characterized in that,
by means of a signaling module (14), control and steering signaling (141 ,142,143) is
generated and transmitted to associated runtime execution modules (30,31 ) via the
runtime interfaces (31 1, 312), wherein the workflow object (71 ,72,73) is processed by
executing the activated process tasks (1921 , 1922, 1923) by means of the runtime
execution modules (30,31 ) based o n the transmitted control and steering signaling
(141 ,142,143), and wherein the process tasks (1921 , 1922, 1923) are generated by
means of process task engines (312, 322).
4. The method according to any one of the claims 1 to 3, characterized in
that the plurality of block elements (1112, 1113, 111 ) is generated based o n
predefined boundary conditions provided by a n automated underwriting system (20,
2 1) by a t least partially extracting data from historical and/or simulated data.
5. The method according to any one of the claims 1 to 4, characterized in
that the automated underwriting system (20, 2 1) is dynamically adapting with regard to
a n overall risk pooled from risk exposed components (50,... 54) by the underwriting
system (20, 2 1) by means of the pattern-recognition based method for a control and
monitoring system (10), by adapting the predefined boundary conditions reflecting the
overall risk appetite of the automated underwriting system (20, 2 1) .
6. The method according to any one of the claims 1 to 5, characterized in
that the one or more of the interconnectable search terms and/or meta-data are
extracted and/or generated at least in part from historical and/or simulated data.
7. The method according to any one of the claims 1 to 6, characterized in
that the block elements ( 1 112, 1113, 1114) comprise, in addition to the one or more of
the interconnectable search terms and/or meta-data, a variable weighting parameter,
wherein each block element ( 1 112, 1113, 1114) and/or its conformity with the stored
block elements ( 1 111, 1112, 1113) and with the recognized block elements with
workflow objects (121 1, 1212) is weighted by the variable weighting parameter for
generating the conformity index.
8. The method according to any one of the claims 1 to 7, characterized in
that the stored workflow objects (121 1, 1212) of the second database (12) are at least
partly generated based on definable boundary condition parameters selecting block
elements ( 1 112, 1113, 1114) from the first database ( 1 1) and/or extracting workflow
objects from historical workflow objects' data.
9. The method according to any one of the claims 1 to 8, characterized in
that the corresponding proximity factor is measured by matching recognized block
elements of the captured workflow object (71 , 72, 73) with block elements (121 1, 1212)
of a workflow object (121 1, 1212) of the second database (12), thereby providing the
measure for the proximity of the two workflow object based o n the mutually
allocatable block elements and/or their degree of conformity.
10. The method according to claim 9, characterized in that, based o n the
filtered block elements of the captured workflow object (71 , 72, 73), a proximity factor
to each of the stored workflow objects (121 1, 1212) of the second database (12) is
determined, wherein a corresponding proximity factor is measured by matching
recognized block elements of the captured workflow object (71 , 72, 73) with block
elements of a workflow object (121 1, 1212) of the second database (12), thereby
providing the measure for the proximity of the two workflow objects based o n the
number of mutually allocatable block elements and/or their degree of conformity.
11. The method according to any one of the claims 1 to 10, characterized in
that the conformity index is generated and assigned to the captured workflow object
(71 , 72, 73) based o n the conformity of the recognized block elements with the stored
block elements ( 1 111, 1112, 1113) of the first searchable data structure ( 1 11) and based
on the conformity of the recognized block elements with workflow objects (121 1, 1212)
of the second database (12), wherein the conformity index provides a measure for the
variance of the captured workflow object (71 , 72, 73) within the predefined boundary
conditions of the automated underwriting system (10).
12. The method according to any one of the claims 1 to 11, characterized in
that the underwriting process flow (19) of the underwriting system (20, 2 1) is based o n
state-transition-controlled processing of workflow objects (71 ,72,73), wherein, by means
of a control system (10), a n object (71 ,72,73) is selected and processed following the
state-structured process flow (19), comprising a plurality of process states
(1911,1912,1913), and, for each process state (1911,1912,1913) , one or more process
tasks (1921 , 1 922,1 923) are executed by means of the control system (10), and wherein
the selected object (71 ,72,73) is processed from one process state (191 1,1912,1913) to
reach a subsequent process state (191 1,1912,1913).
13. The method according to any one of the claims 1 to 12, characterized in
that state parameters of a workflow object (71 ,72,73) are captured by capturing means
(15,151 ,152,153,1 54) of the control system (10), and a process state (191 1,1912,1913) is
determined based on the captured state parameters and assigned to the selected
object (71 ,72,73), wherein, based o n the determined process state (191 1,1912,1913)
and/or state parameters of the workflow object (71 ,72,73), a t least one process task
(1921 , 1922, 1923) is generated by means of the monitoring and control system (10) for
a specific process state (1911,1912,1913) .
14. The method according to claim 13, characterized in that a generated
process task (191 1, 1921 , 1923) is activated a s a function of task parameters assigned to
a process task (1921 , 1922, 1923).
15. The method according to any one of the claims 1 to 14, characterized in
that the process flow (19) is dynamically operated by the control system (10), wherein,
by means of the control system (10), a n underwriting workflow object (71 ,72,73) is
processed from the determined process state (121 ,122,123) to reach a subsequent
process state (191 1,1912,1913) by executing the assigned process tasks (1921 , 1922,
1923, 1924).
1 . The method according to any one of the claims 1 to 15, characterized in
that the underwriting process flow (19) is dynamically generated and adapted, wherein
the workflow object (71 ,72,73) is processed by means of the control system (10) that
initiates a subsequent process state (191 1,1922,1933) by triggering defined trigger
values of the task parameters and/or state parameters of the preceding process state
(191 1,1922,1933).
17. The method according to any one of the claims 1 to 1 , characterized in
that, for the subsequent state transition within the process flow (19), the subsequent
process task (191 1, 1912, 1923) is split into subtasks, wherein a subtask is generated by
the control system (10) in order to provide the additional process states triggered and
assigned to the process flow (19), when a measurement of the conformity index is taken
that exceeds the risk threshold value (141 ) , a s established by means of the trigger
module (14).
18. The method according to any one of the claims 1 to 17, characterized in
that the control system (10) comprises a historical engine device ( 1 6), wherein historical
data of processed workflow objects associated with known risk parameters are stored
in a storing device of the control system (10), wherein the stored historical data are
compared to the present workflow object (71 , 72, 73), and relevant historical workflow
object (71 , 72, 73) data are filtered from the stored data by means of a filter-module
( 1 7), wherein the historical engine device ( 1 ) and the control system (10) are
connected by a data link for data signaling transmission between the control system
(10) and the historical engine device ( 1 6), and wherein the state-structured process
flow (19) is dynamically generated by the control system (10), and the captured
workflow object is dynamically processed based o n the data signaling transmission
from the historical engine device ( 1 ) and the process management engine (13) of the
control system (10).
19. The method according to any one of the claims 1 to 18, characterized in
that the control system (10) and the runtime execution devices (30,31 ) interact in
runtime, wherein the object (71 ,72,73) is processed based o n the dynamically adapted
process flow (19) with the generated process tasks (1921 , 1922, 1923) by executing the
activated process tasks (1921 , 1922, 1923) by means of the runtime execution devices
(30,31 ) based o n the transmitted control and steering signaling (141 ,142,143).
20. The method according to any one of the claims 1 to 19, characterized in
that a first-tier execution of the control system (10) comprises the generation of the
process states (191 1,1912,1913), and a second-tier execution of the control system (10)
comprises the generation and association of the additional process tasks
(191 1,1912,1913), when a measurement of the conformity index is taken that exceeds
the risk threshold value (141 ) , a s established by means of the trigger module (14),
wherein the processing of the workflow object (71 ,72,73) and the appropriate signaling
(141 ,142,143) by the control system (10) is dynamically adapted by alternating
operating parameters of the runtime execution devices (30,31 ) .
2 1. The method according to any one of the claims 1 to 20, characterized in
that the control system (10) is self-adapting by automatically capturing processed
workflow objects (71 , 72, 73) after finishing the processing by means of the underwriting
process flow (19), wherein the content of the second database (12) is dynamically
altered storing the captured and processed workflow object (121 1, 121 2) by means of
the second searchable data structure (121 ) of the second database (12).
22. The method according to any one of the claims 1 to 2 1, characterized in
that the capturing means (15) comprises measuring devices and/or measuring sensors
(151 ,152,153,154) for capturing the state parameters and/or task parameters of the
workflow object (71 ,72,73).
23. The method according to any one of the claims 1 to 22, characterized in
that the parameters of the workflow objects (71 , 72, 73) are captured via a plurality of
input devices (181 ,182,183) that are accessible by risk exposed components (51 ,52,53)
and/or automated insurance systems (20,21 ) for dynamically adapting the workflow
objects (71 ,72,73) via the interface module (18) and/or network (60) based o n
appropriate signal generation by means of the signaling module (14) of the control
system (10).
24. The method according to any one of the claims 1 to 23, characterized in
that at least parts of task parameters of the process tasks ( 1921 , 1922, 1923) of the
additional process states (191 1, 1912, 1913) are dynamically generated by the control
system (10) based o n the measured conformity index of the captured workflow object
(71 ,72,73).
25. The method according to any one of the claims 1 to 24, characterized in
that appropriate signaling (141 ,142,143) is generated by means of the signaling module
(14) for steering the execution devices (50,51 ,52) that process the selected object
(71 ,72,73) according to the process flow and/or additional process flow (19) a s
generated by the control system (10).
25. The method according to any one of the claims 1 to 24, characterized in
that the processing of the captured workflow object (71 ,72,73) is monitored by means
of dedicated monitoring and/or measuring devices (30, 3 1) of the control system (10)
based a t least a n the captured state parameters, the tasks parameters and/or
operating parameters.
26. The method according to any one of the claims 1 to 25, characterized in
that by means of the additional process states, which are triggered and assigned to the
process flow (19) by means of the process management engine (13) of the
measurement and monitoring system (10), the workflow object (71 , 72, 73)
automatically is assembled within the workflow pathway (19) by means of the system
(10) based o n a predefined structure by means of composing a plurality of block
elements ( 1 112, 1113, 1114) triggered within the first database ( 1 1) .
27. The method according to any one of the claims 1 to 26, characterized in
that by means of the additional process states, which are triggered and assigned to the
process flow ( 19) by means of the process management engine (13) of the
measurement and monitoring system (10), based o n the conformity of the recognized
block elements with workflow objects (121 1, 1212) of the second database (12), a t least
one risk-importing recognized block element of the workflow object (71 , 72, 73) is
substituted by block elements ( 1 112, 1113, 1114) triggered within the first database ( 1 1)
improving the measured conformity index.
28. The method according to any one of the claim 27, characterized in that
risk-importing recognized block elements of the workflow object (71 , 72, 73) are
substituted by block elements ( 1 112, 1113, 1114) triggered within the first database ( 1 1)
until the measured conformity index of the captured workflow object (71 , 72, 73) is
measured within the predefined boundary conditions of the automated underwriting
system (10).
29. A pattern-recognition based control and monitoring system (10) for
automated underwriting systems (20, 2 1) or automated pattern-recognition based
data-capturing systems, wherein a portfolio of risk transfer objects of the underwriting
system (20, 2 1) is monitored by consecutively capturing risk transfer objects (71 , 72, 73)
of the portfolio of pooled risk-transfer objects (71 , 72, 73), and wherein a n activation of
process states (191 1,1912,1913) is triggered upon detecting non-conforming risk-transfer
objects (71 , 72, 73) within the portfolio of pooled risk-transfer objects (71 , 72, 73) by
means of the control and monitoring system (10), characterized,
in that a plurality of block elements ( 1 112, 1113, 1114) are generatable
based o n predefined boundary conditions provided by means of a n automated
underwriting system (20, 2 1) , wherein the block elements ( 1 112, 1113, 1114) are
triggerable parts of risk-transfer objects (71 , 72, 73) and comprise one or more of the
interconnectable search terms and/or meta-data, and wherein the block elements
( 1 112, 1113, 1114) are assigned to a recognition block-map ( 1 11) and stored by means
of a first searchable data structure ( 1 12) provided by a first database ( 1 1) ,
in that the system (10) comprises a second database (12) that provides a
second searchable data structure (121 ) for storing a plurality of risk-transfer objects
( 1 2 11, 1212), wherein the stored risk-transfer objects ( 12 11, 12 12) of the second
database (12) are at least in part generated based o n definable boundary condition
parameters and/or generated based on historical risk-transfer objects,
in that a risk-transfer object (71 , 72, 73) is capturable within the workflow
pathway (19) by means of the measurement and monitoring system (10), wherein the
system (10) comprises a core engine (131 ) with a recognition module (132) for scanning
the captured risk-transfer object (71 , 72, 73), thereby recognizing and identifying block
elements ( 1 112, 1113, 1114) of the stored recognition block-map ( 1 11) in the captured
risk-transfer object (71 , 72, 73),
in that, based o n the filtered block elements of the captured risk-transfer
object (71 , 72, 73), a proximity factor with regard to each of the stored risk-transfer
objects (121 1, 1212) of the second database (12) is determinable, wherein a
corresponding proximity factor is measured by matching recognized block elements of
the captured risk-transfer object (71 , 72, 73) with block elements of a risk-transfer object
(121 1, 1212) of the second database ( 1 2), thereby providing the measure for the
proximity of the two risk-transfer object based o n the mutually allocatable block
elements,
in that a conformity index is generatable and assignable to the captured
risk-transfer object (71 , 72, 73), based o n the conformity of the recognized block
elements with the stored block elements (1111, 1112, 1113) of the first searchable data
structure ( 1 11) and based o n the conformity of the recognized block elements with risktransfer
objects (121 1, 121 2) of the second database (12), wherein only risk-transfer
objects ( 1 2 11, 1212) within a predefined range of the proximity factor are selected for
generating the conformity index, and wherein the conformity index provides a measure
for the conformity of the captured risk-transfer object (71 , 72, 73) within the predefined
boundary conditions of the automated underwriting system (10), and
in that the process states are triggerable and assignable to the process flow
(19) by means of a process management engine (13) of the measurement and
monitoring system (10), when a measurement of the conformity index is taken that
exceeds a risk threshold value (141 ) , a s established by means of the trigger module (14).
30. The system according to claim 29, characterized in that, for recognizing
and identifying block elements ( 1 12, 113, 114) by means of the recognition module
(132), the captured workflow object is content-scannable by means of the recognition
module (132) triggering for block elements of the workflow objects (71 , 72, 73) by
means of the stored block elements of the first searchable data structure ( 1 11) , wherein
triggered block elements of the captured workflow object are classified by means of a
classification module (133), wherein associated entities are extracted by means of a
entity extraction module (134), and wherein detected search terms of the workflow
object are flagged by means of a key-word flagging module (135).