Article Review: Case Analytics Workbench for Hybrid Process Modeling

This report provides a comprehensive review of the article "Case Analytics Workbench: Platform for Hybrid Process Model Creation and Evaluation." The article explores the limitations of declarative and imperative processes and proposes a hybrid approach to business process modeling. It introduces the Case Analytics Workbench as an end-to-end business process system, leveraging both declarative and imperative modeling techniques. The review highlights the workbench's modules, including Data Management, Process Mining, Evidence Management, and Case Model Management. It analyzes two case studies from the healthcare and insurance industries, demonstrating the workbench's effectiveness. The report also discusses the strengths of the hybrid approach, such as support for customized modeling and user interaction, while acknowledging limitations like the need for empirical evaluation and the complexity of the data management module. The conclusion emphasizes the need for further research, particularly in the integration of declarative models and the handling of temporal constraints.

Business Process Engineering
BPM Article Review
ARTICLE REVIEW
OF
CASE ANALYTICS WORKBENCH:
PLATFORM FOR HYBRID PROCESS MODEL CREATION AND EVALUATION
Authors: Yiqin Yu, Xiang Li, Haifeng Liu, Jing-Mei, Nirmal Mukhi, Vatche Ishakian,
Guotong Xie, Geetika T. Lakshmanan, and Mike Martin
Prepared by:
Client’s Name
Student’s ID:

Process 2
Introduction
The article on workbench analytics regarding the process of hybrid model evolution and
creation is taken from the occurrence of the International Conference process. It produces an
improvement that discusses the modeling in the business process. Recent alternatives in the
declarative and imperative processes indicate general limitations. Hence, combining the
mentioned processes to come up with a hybrid process takes advantage of the various
processes and seems as the better alternative in modeling. In addition, these authors provide
an assistance relevant to modelers during creation and refining of the model which is a hybrid
process. The process of modeling produced workbench case analytics that was considered to
be an end-to-end business process system. Also, the two study cases that involved health care
and insurance industry were analyzed to present deeper understanding regarding the model in
discussion (Lano, et al., 2014). The paper at hand produces a summary of the article that
includes the weaknesses and strengths of the declarative and imperative approaches. The
reasons for the hybrid process model developed an attractive process of business that
combined both approaches in business. In the strategy of providing evidence of the idea,
articles that were related to the process were compared. A focus on the case studies illustrated
the applicability and effectiveness of the workbench case analytics that applied in reality.
Lastly, a limitation was in the model tool workbench that was to be considered during further
application that is to be presented in this article (Scheer, 2013).
Body
Imperative declarative and modeling approaches give critical information on the management
business process domain. The former has a structured model for modeling management in
business process as opposed to the later that supports languages that require flexibility. Also,
there is a demonstration in the imperative model specification that specifies how the whole
process is being done. The present declarative approaches have an eye on the logic that
determines the actions in play during process description in the activities being performed
and the constraints that prohibit the undesired behavior. These imperative features can be
proved in the research program that delivers sequential information that is better. The
declarative programs gives an insight on the circumstantial information. The process includes
imperative models that are the BPEL, XPDL, and the BPMN together with the declarative
models that include the DCR Graphs, Declare and the SCIFF that show the benefits of
process modeling (Lano, et al., 2014).
However, the study indicates that the entire business process modeling that is in practice is
most likely to be impossible when used in declarative models since there are processes or any

Process 3
relevant parts in the process that could be present in a better imperative model. Examples
include the Business Process Modelling Notation together with the Event-Driven Process
Chains. These two modeling imperative techniques produce the execution activity sequence
explicitly whose result heavily depend on the workers’ knowledge, over specification and
inexistence of flexibility. In addition, the declarative modeling tool takes Declare as an event
that is an ‘outside-in’ strategy describing the essential characteristics other than the working
process in whole (RADHAKRISHNAN & BALASUBRAMANIAN, 2008).
Taking reference from the shortcomings and the advantages that come from the two types in
modeling business process, the known authors argue that overall, the approach in hybrid
modeling seems to be more attractive in the business modeling process. These authors show
their evidence on the related work about the models of the hybrid process that already exists
and have been developed in academic and industrial works. The case manager of IBM (ICM)
is an example of designing the major model structure together with the process task using the
Case Management Model and Notation – the declarative language in modelling; and the
Business Process Model and Notation – the imperative language in modelling respectively.
The authors also put down research that gave a suggestion that produced better business
modeling process during the division of event logs into the unstructured and the structured
sequences that discovered the imperative sub-processes together with the declarative sub-
processes in the hybrid mining process approach. Also, there was an introduction of
integrating methodology in the imperative and the declarative modeling processes whose
approach is used in forming model business process in the process of logistic service. Its
application was in the Chinese bulk port. The case study provides proof of the effectiveness
in loose-structure modeling process service together with the highly structured parts. Other
than the modeling process supporting the corresponding rules and logistics regulations that
are put across in various organizations, the process also gives an assistance in strict order
execution land the activity process in the specific organization to be done completely (Neiger,
et al., 2008).
Taking from the terms in the Case Analytics Workbench, authors state that this system is end-
to-end including the various advantages that are beyond the two modeling approach
combination. To begin with, the approach can provide support in customized modeling
process by providing leverage of clustering event log. Also, rich users can interact and
visualize; as a result, the evolution process and the modeling creation can be facilitated.
Lastly, there is an integration with the ICM having to be very useful. Case Analytics
Workbench is present in industries such as healthcare and insurance and shows how it has its

Process 4
importance in the modeler during the creation of model processes inconvenient and less
complex methodology (Lano, et al., 2014).
The Case Analytics Workbench has a division of four major modules; Evidence
Management, Case Model Management, Data Management and Process Mining Module. It
also provides a separation into two parts such as the client-side together with the server-side
components. As the former gets deployed in the big cloud environment, the other helps in
rendering user interfacing.
The Data Management is useful in dealing with a number of processes in execution logs that
have a transformation in the formal format of event logs after which it is processed using data
clustering. Next is the event logs coming from the Data Management that is executed during
the process Mining Module whose combination of imperative and declarative mining
processes are run. The resulting produce in this mentioned step is the Evidence Adapter 1
together with the Evidence Adapter 2. The stage that follows is the Evidence Management
Collection of the Adapter and giving them to the user for better understanding and sight.
Lastly, the Case Model Management includes the Case Model Manager, Case Model Editor,
and the Case Model DB that provides the case model that can be used by a user in interacting
with the model creation, saving, editing, and transformation.
Case Analytics Workbench effectiveness is proven using the insurance company case study.
During Data management stage, the 4305 execution logs that were the process of
underwriting were changed to formal logs of the event by using Data Preparation. The change
led to the generation of two clusters that include the 2038 case instances that related to ‘the
auto quote’ together with the 2267 case about ‘property quote’. Users can henceforth easily
determine the cluster 1 and cluster 0 that represent the cases in auto insurance also the
property insurance cases respectively. A change to the known Process Mining Module that is
a method of the declaration, has the goal ‘cycle time being two days less’ (Neiger, et al.,
2008).
The declaration process–mining engine produced the additional constraints and the additional
tasks that are not visible in the fresh model that is raw. The results are represented
appropriately with more details put in the support degree. Also, the goal achievement
correlation can be added. The stage that follows is the Evidence Management that involves
the imperative and declarative mining processes in the engines used. Evidence that possesses
the support figure that is greater than 0.3 and has a greater correlation goal than 0.5 can be
filtered off.

Process 5
The usability of Case Analytics Workbench can be put down like the case study in the Care
Pathway Refinement. The present authors the possibility of managing the congestive failure
in the heart due to using workbench that is complex with string requirement flexibility for ad
hoc variation. The tests in the traditional baseline are to be done in the care pathway before
starting any treatment. However, the original Electronic Medical Records information model
of pathway care containing numerous baseline tests had not been done but the treatments
were started still. The reasons for such kind of violation are reasonable. Also, the treatments
were able to be done in accordance with the previous medical histories that could miss in the
record, thereby initiating treatments can be performed with lack of past baseline test in few
cases (Jack, et al., 2012).
When discussing and concluding, the authors in discussion produced two viewpoints that
were comprehensive, one originated from system design whereas the other originated from
applicability. The last view indicates the component cost that includes the Data Clustering
decoupling that does exist. Moreover, there could disadvantages when a large volume of
information required in transmission in Data Management Module together with various
engines in mining process since the mentioned Data Management Module provide
component independence. The last view indicates that the positive feedback from
professionals that are related that include product managers in BPM and the clinical
physicians and the developers. Such feedbacks come from the advantages existing in
combining declarative and imperative process model in the Case analytics Workbench since
user interactions, interaction mode design, and the analytic result visualization (Scheer,
2013).
Conclusion.
The mentioned authors produced the very common limitation that is spread in the research.
These authors state that there is a necessity in getting a better empirical evaluation produced
by various modelers. The generated models are better evaluated and require transformation
and exportation to various platforms in management business process that include the ICM
since when used as a modeling tool, the workbench will not perform well. There also is a
clear demonstration of more points that limit the research that was done. A hybrid model
process has to be built with regards to transparent model element separation; in consequence,
Task and Stage elements are to be considered as Step elements and declarative parts in the
Process Task in imperative ones. In the contrary, most of these cases in an actual sense are
difficult to identify the elements of modeling. Hence, such cases a hybrid model process has
to be designed with flexibility and considering both the imperative and declarative models. A

Process 6
suggestion for further consideration is also put forward in that the getting more branches of
temporal constraints and data conditions, the declarative portion of the model should
integrate Declare other than using a GSM-based model whose support is limited (Scheer,
2013).

Process 7
References
Jack, D., Gulledge, T. & Lee, C., 2012. Business Process Engineering: Advancing the State
of the Art. illustrated ed. Mackay: Springer Science & Business Media.
Lano, K., Singh, R. & Maroukian, K., 2014. Model-Driven Business Process Engineering.
reprint ed. Darwin: Bentham Science Publishers.
Neiger, D., Churilov, L. & Flitman, A., 2008. Value-Focused Business Process Engineering :
a Systems Approach: with Applications to Human Resource Management. illustrated ed.
Sydney: Springer Science & Business Media.
RADHAKRISHNAN, R. & BALASUBRAMANIAN, S., 2008. BUSINESS PROCESS
REENGINEERING: Text and Cases. 1 ed. Melbourne: PHI Learning Pvt. Ltd.
Scheer, A., 2013. Business Process Engineering Study Edition: Reference Models for
Industrial Enterprises. illustrated ed. Darwin: Springer Science & Business Media.