Data-Driven Performance Analysis of Scheduled Processes in Business Process Engineering

This essay reviews the issues of data-driven performance analysis for some scheduled process. The authors have developed a technique that involves or amalgamates the techniques from Queuing Theory with a Coloured Petri Nets. Particularly, for the purpose of effectiveness of computation, the study has defined a state of folding operation that give the permission to project creation of QCSPN model into the Queuing network.

Running head: BUSINESS PROCESS ENGINEERING
Business Process Engineering
Name of the Student
Name of the University
Author Note

1BUSINESS PROCESS ENGINEERING
Topic- Data-Driven Performance Analysis of Scheduled Processes
Introduction
This essay is performed on the “13tth International Conference on Business
Process Management (BPM)”. The essay reviews the article: “Data-Driven Performance
Analysis of Scheduled Processes” which is based on Business Process Management. The
major purpose of the essay is to understand facts and findings on Business Process
Management and its use in the contemporary business. In order to support the review, three
peer-reviewed journal article have been selected that have been performed on Business
Process Management. In order to review the journal article, essay paid head to all sections of
the article starting from the review of the introductory part, followed by literature review and
methodology. When reviewing the paper, the essay first introduces how the topic in the paper
has been introduced and what goals and objectives have been prepared. Likewise, the essay
also focuses on the research techniques that have been used by the author of the article. When
discussing the methods, the essay also identifies the reliability and validity of the methods by
assessing the methods used in other similar articles. The major part of the review is presented
on the basis of the comparison between the conclusions drawn by the authors and conclusions
drawn by other articles on the selected context. Moreover, the methods used in the article
have been compared to the methods used in supporting articles performed on Business
Process Management.
Analysis
The major intent of the article is to focus on the issue related to the performance of
schedule business processes which is a fundamental significance of service as well as the
manufacturing technique. This happened as the existing method for performance analysis is
not able to deal with both queuing semantics as well as the process perspective. Therefore,

2BUSINESS PROCESS ENGINEERING
the paper particularly aims to resolve this gap by enhancing a significant method for
implementing rich process logs to particularly analyse the performance of scheduled process.
The issue highlighted in this paper is a crucial, as the study performed by Giacosa,
Mazzoleni and Usai (2018) also states that the business process management is always a
critical factor as most of the small and medium size firms tend to adopt process organizations
but hardly a business pay head to the process by which organizations can remain distinctive.
On the other side, a study performed by Malinova and Mendling (2018) reveals the fact that
actionable guidelines are missing from several reference that particularly works on the
Business Process Management. The authors claims that the success factors are not properly
contextualized in several sections or in the element of the business process management.
So, in order to develop a unique and core method for implementing the rich process
logs to assess the performance of the scheduled process, the study uses the method that clubs
simulation, queuing analytics as well as the statistical techniques. The fundamental approach
of the method is the discovery of an individual case model from data on the basis an
extension of Coloured Petri Nets Formalism. It is worth mentioning that the resulting model
discussed in the study can help to stimulate to respond to the performance queries (Miglietta
et al. 2018). When particularly applying the methods, the authors have discussed practical
approach in which they referred a use-case that is supported by the data from Dana-Farber
Cancer Institute. The scholars particularly focus on the fact of schedule of an ambulatory
patient which typically includes a blood draw – an experimentation by a physician or the
nurse and practitioner as well as a chemotherapy infusion. In the study, the authors have
talked about a set of models that such a schedule log which particularly serves as the data
model as well as Coloured Petri Nets which is a modelling formalism based on a Coloured
Petri Nets with stochastic delays as well as scheduling transitions. On the contrary, the
schedule log includes a series of tasks and the particular execution of time of performing the

3BUSINESS PROCESS ENGINEERING
tasks. According to Giacosa, Mazzoleni and Usai (2018) the significance of developing
targets at the time of modelling techniques should be emphasized, particularly during the
identification of techniques that could undergo the phase of BPM lifecycle. In addition, the
authors of this article had also mentioned that having developed goals could help to enable
the organizations or the businesses to measure the performance of the process after it has
gone through the redesign.
Another significant section found in the article is discovery of queue-enabling CSPN
models. This particular section of the article is devoted to the discovery as well as enrichment
of Queue-Enabling Colored Stochastic Petri Nets which is derived from the scheduled log.
When implementing this approach, the author have provided an overview of pre-processing
as well as the assumptions necessary for finding out and enriching the QCSPN model and
thereafter the stud demonstrate a three-step discovery algorithm that usually constructs
QCSPN. According to Malinova and Mendling (2018), precedence constraints are the key
feature of scheduled process which helps to make sure that cases are not authorized to
continue to a new task before the subset of other related activities are performed.
Particularly, in order to deal with the parallelism, the authors of the study have
applied the method of pre-processing phase under which they have depicted parallel task
activities with the help interval calculus. Thus, the study assume the activities to be parallel if
the intersection of their planned time is not empty. However, Miglietta et al. (2018)
mentioned that the study needs to assume the existence of parallelism set II, which further
contains a series of activities or tasks that are planned to be performed in parallel. Moreover,
this particular discovery process helps to derive three more assumption such as work
conserving, temporal deviation and duration dependencies. As put forward by Malinova and
Mendling (2018), the resources could be available immediately after finishing a particular

4BUSINESS PROCESS ENGINEERING
activity, while the temporal deviation could be of the schedule in which the activities could
deviate in a time period with the inclusion of resource, activities as well as routing deviation.
Apart from the above presented section, the article also presented a set of facts related
to Folding and Projection of QSCPN particularly into Queuing network. This section of the
paper helps to observe the fact that authors have emphasized on the folding and projection of
QCSPN. However, Giacosa, Mazzoleni and Usai (2018) commented that the authors have
only discussed the state of the function but the definition lack required details about the
function should be implemented. According to the authors of the selected article, a folding
function only creates a new QCSPN model under a set of assumption. For example, removed
parallelism in folding 1 is particularly known for its negative impact on the analytical
tractability of queuing network. This could particularly motivates people to consider a folding
operation which converts N0 into a concurrency-free model. This particular function helps to
enable all parallel task to be initiated as well as ended at the same time covering all necessary
resources that were later scheduled to carry out the parallel activities.
As put forward by Malinova and Mendling (2018), the institution behind the
perspective of QCSPN into Queuing Network remains straight forward; moreover queuing
network must include a directed graph, with vertices being single-station queue as well as
edge. Nonetheless, the article did not mention about the fact that a first step of projection
queuing stations of generating QCSPN are converted into vertices which belong to target
Queuing Network. According to Miglietta et al. (2018), due to this stated reason every vertex
of the Queuing network could be characterized by a stimulated number of resources that exist
in that particular vertices.
The evaluation section of the article has discussed the outcome of an empirical
evaluation of the proposed approach. Hence, the authors of the articles have mentioned about

5BUSINESS PROCESS ENGINEERING
and demonstrated the relevancy of the method spectrum on the basis of Petri Net-based
simulators with the help of version of the originating model. The scholars particularly focus
on the fundamental aspects that further include the execution of their used techniques. The
study also discussed the dataset and a complete design of their experiment. When it comes to
the implementation, the article has implemented the model construction, its enrichment with
the past data findings, the stimulation semantics and the folding operation along with a
projection into the Queuing Theory on the basis of the snapshot predictor. The
implementation has been done with the help of python programming language which helps to
build an open-sources SNAKES model. It has been identified that the implementation is
available and it remains as the free-open source project. Another significant part of the
conclusion drawn in the article is the data description; this means that the data that authors
consider usually come from Dana-Farber Cancer Institute. The study has combined two
different dataset into a single one which includes schedule visits as well as their
corresponding execution times. Hence, the later one contains a detailed schedule for each day
while the other one is particularly based on Real Time Location System.
In addition to this, the empirical evaluation also demonstrates that efficient queuing
predictor could remain more accurate when it is particularly compared to stimulation model.
According to the findings, the major drawback of the projected model is that it might not be
applicable to the queries of performance question. For instance, the prediction of individual
resources utilization, without the using the shared resources as well as parallelism prediction.
It is also found that Classical Queuing Networks are not particularly expressive enough to
make analysis of such questions without stimulating. In addition, Giacosa, Mazzoleni and
Usai (2018) commend that folding parallelism as well as the shared resources did not have
impact on the model of stimulation and this should be explained by the fact that the dataset
presented in the article includes some parallel tasks. These tasks are required to be implanted

6BUSINESS PROCESS ENGINEERING
by a single resources. When reviewing the articles, Malinova and Mendling (2018) also
mentioned that the major error that is found in the finding is the additional insight. According
to this author as the queuing predictor develops upon the Rescheduled folding, it could avoid
scheduling delays and carries major negative error in the eventual result.
Despite the criticism of the authors, a significant factor, which is worth mentioning is
that deviation in the order of tasks are not rate and such nature could explain then inaccuracy
of the stimulation of the models as they consider that the sequence of takes are not destroyed.
Nonetheless, the queuing predictor might consider a series of activities or task despite their
executional order, which further describes their accuracy. However, it is worth mentioning
that the findings of the selected paper lack an enhanced integrated framework for the
Business Process Management with some experimental guidance related to each of its
elements.
Conclusion
This essay reviews the issues of data-driven performance analysis for some scheduled
process. In order to finalize this, the authors have developed a technique that involves or
amalgamates the techniques from Queuing Theory with a Coloured Petri Nets. Particularly,
for the purpose of effectiveness of computation, the study has defined a state of folding
operation that give the permission to project creation of QCSPN model into the Queuing
network. The approach in the article was carried out and evaluated with the help of
experimental and practical data from outpatient cancer institutions demonstrating the
influence of model abstraction on transparency with respect to root-mean squared error. The
essay particularly focuses on the issue that existing technique for performance analysis is not
able to consider queuing semantic and the technique of different perspective. The essay
reviews how this gap has been covered in the paper with variety of methods.

7BUSINESS PROCESS ENGINEERING

8BUSINESS PROCESS ENGINEERING
Reference-
Giacosa, E., Mazzoleni, A. and Usai, A., 2018. Business Process Management (BPM) How
complementary BPM capabilities can build an ambidextrous state in business process
activities of family firms. Business Process Management Journal. 3 September 2018,
24(5):1145-1162
Malinova, M. and Mendling, J., 2018. Identifying do’s and don’ts using the integrated
business process management framework. Business Process Management Journal. 2 July
2018, 24(4):882-899
Miglietta, N., Battisti, E., Carayannis, E. and Salvi, A., 2018. Capital structure and business
process management: evidence from ambidextrous organizations. Business Process
Management Journal. Vol. 24 Issue: 5, pp.1255-1270, https://doi-
org.wallaby.vu.edu.au:4433/10.1108/BPMJ-07-2017-0214
Senderovich, A., Rogge-Solti, A., Gal, A., Mendling, J., Mandelbaum, A., Kadish, S. and
Bunnell, C.A., 2015, August. Data-driven performance analysis of scheduled processes. In
International Conference on Business Process Management (pp. 35-52). Springer, Cham.

9BUSINESS PROCESS ENGINEERING