Contextual Representations for Enterprise Model Application

(C.R.E.M.A.)

Nikolaus Wintrich and Malte Meißner

Corporate Management, Fraunhofer IPK, Pascalstr. 8-9, Berlin 10587, Germany

Keywords: Contextual Perspectives, Enterprise Models, Business Process Models, Users, Complexity, IEM, Process

Assistant.

Abstract: Enterprise models help to make enterprise processes and services more transparent while also improving

stockholders’ abilities to analyze, optimize and control them. These models include every relevant element

as well as enterprise processes, including their relationships with one another. the growing complexity of

these models reduces their usability as well as user friendliness. the different stakeholders have different

expectations towards the visualization and the relevant information being displayed to them. Therefore

contextual representations of enterprise models allow the provision of the relevant information in a user

oriented perspective by applying different visualization methods. This helps to manage the complexity and

facilitates a sustainable application of enterprise models for various enterprise aspects such as quality

management, strategic planning, reporting and operational control.

1 INTRODUCTION

Enterprise modelling can be defined as the

systematical description of all the elements, business

processes and the relationships amongst one another,

which are relevant to a certain investigation

objective (Schwermer, 1998). An enterprise model

includes the structure, behavior and organization of

the enterprise in order to provide a general

understanding (Vernadat, 2002).

The application of enterprise models has

increased dramatically over the recent years and

their role is to support several business activities like

business process management, quality management,

production planning, strategic planning and resource

management. This results in an exponentially

increased size of such models to fulfill the

requirements of describing the whole enterprise,

including its processes, order information, products

and resources like human resources, documents, IT

systems, machines as well as equipment.

Another driver for the increased size is the

overall approach to integrate information from

different domains into one information backbone

(e.g. product information - typically contained in a

PLM system with processes – typically contained in

an enterprise or business process model) or at least

to interlink these information elements so that

dependencies can be identified and described.

The increased size of enterprise models along

with the approach of higher information integration

results into an increased model complexity. On the

one hand, this makes it more difficult for the

modeler to maintain and update the model. On the

other hand all stakeholders which are applying these

models are facing the problem of fast and easy

information retrieval. They are especially interested

in extracting concrete information. In terms of

enterprise models they are specifically interested in

some elements like concrete production processes,

quality related documents or they are looking for

detailed information about the involvement of a

single role within the whole enterprise. to put this

into numbers: current versions of our customers

enterprise models have more than 15.000 elements

and consist of more than 220 partial processes (up to

6 levels). These already huge numbers are going to

increase rapidly within the future. We’re expecting

to have enterprise models consisting of up to 50.000

elements and 500 partial processes within the next

two years.

to meet the demands from our customers to offer

methods and possibilities as well as tools to manage

and apply such huge enterprise models in the future

we present a framework for contextual model

representations, which allow us to compile a suitable

483

Wintrich N. and Meißner M..

Contextual Representations for Enterprise Model Application (C.R.E.M.A.).

DOI: 10.5220/0005243804830489

In Proceedings of the 3rd International Conference on Model-Driven Engineering and Software Development (MODELSWARD-2015), pages 483-489

ISBN: 978-989-758-083-3

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

and applicable representation of the necessary model

elements for the individual tasks of the stakeholders.

Since the information types within an enterprise

model are very heterogeneous we believe that a

sustainable and practical usage can only be achieved

by offering a wide set of representations and

utilizing different visualization techniques for the

respective information types. The rest of the paper is

organized, so that in section 2, we will establish

stakeholder needs and demands by looking into

several studies. Section 3 will see us using these

criteria to evaluate related research. In section 4 we

will introduce our framework for contextual

enterprise model representations. This will include

an application of said framework. The conclusion

will be stated in section 5.

2 REQUIREMENTS AND

CHALLENGES

In order to satisfy stakeholder demands, it is

important to identify these demands. Several studies

have dealt with Business Process Management

(BPM) and their results help us to identify the main

requirements for BPM-systems. These requirements

can be directly translated to enterprise models since

they are ever increasing in importance. The

following studies deal with certain aspects and

criteria for BPM models, as well as enterprise

models. At the end of this section, we will have

identified five core requirements through the

analysis of these studies.

The first study deals with the criteria themselves,

in particular, how study participants define

understandability. The study „Understanding

Understandability of Conceptual Models – What Are

We Actually Talking about?“ (Houy et al., 2012)

shows a diverse collection of criteria and argues for

a more unified approach to measuring

understandability for which it offers some

guidelines. The scope of this study does not allow

for an in-depth analysis of the different criteria, but

it is important to notice that different studies employ

different criteria and degrees of understandability

when they look at business- and process-models.

This study thus serves as a cautious reminder that

understandability can be an ambiguous term.

Bobrik has defined three foundational principles

which can improve the understandability of process-

visualizations (Bobrik, 2008). These principles can

easily be projected onto business process models.

the first principle is the “Notation“ of symbols:

different shapes and colors can enhance user

understandability. Secondly, the “Layout”, the

organization and structure of elements within the

model is identified as essential. The last principle is

“Reduktion“(reduction) which deals with the

aggregation of model elements. These principles

can all be found in one way or another in each of

these studies, as well as subsequent sections.

The next study focusses on understandability as

well as how participants rate their own capabilities

of understanding a process-model (Mendling et al.,

2007). The study, conducted among 73 students of

the field process management and several experts,

offers two important results: personal characteristics

as well as the size of the model (its complexity), are

the two main factors for model understandability.

Both results are equally apparent for students and

experts of process models. Since this study is from

2007, we can assume that the process models that

were used then, would be even larger now, making

the size even more important. the size and

complexity of models is a reoccurring challenge in

every study and most research papers that deal with

enterprise models and process management. It is

thus a core stakeholder demand. Also, the individual

preferences as an understandability criterion is

important, as this will be addressed later in section 4.

The growing complexity of business and process

models is also identified as the major issue by a

study by Bearing Point (BARC, 2013). along with

the growing complexity of the employees’

responsibilities, this study reveals that most

companies use several process-models for different

processes. a coherent, diverse, context sensitive

business model would connect these different

processes as well as save resources, thus we can

again identify complexity reduction as a major

requirement for a contextual representation of an

enterprise model. Other aspects that are mentioned

in this study are forecasting, making budget

decisions and planning ahead. These aspects are

reoccurring and will be collected under the umbrella

term transparency as a requirement.

the last two studies offer two essential set of

results, with the first one focusing on companies and

the second one on individuals. the former

(BearingPoint 2012) shows that the general

importance and usefulness of sustainable BPM is

accepted, yet there are still obstacles for companies

to integrate process management. besides reluctance

from management, the study identifies model

complexity and the missing holistic nature of many

BPM models as the main challenges. It can be

assumed that, if the two aspects mentioned will be

MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment

484

improved, management would tend to be more

supportive. The majority of companies stated that

efficiency and standardization are their top demands

for their BPM systems. These are also requirements

we can apply to an enterprise model representation.

in the last study (Harmon and Wolf, 2014),

which is also the largest we look at, we see that an

increase in efficiency is the number one priority for

most BPM user. Even more importantly though is

the lack of innovation in the BPM world. The needs

and demands have largely remained the same over

the past 8 years. The study shows that the market is

growing “slow and steadily”, yet it is important to

note, that many issues largely remain the same. So,

the requirement we draw out of this study is the need

for a more efficient process management, along with

a unified, holistic enterprise model for the entirety of

an enterprise. Furthermore, this study shows that the

market is ready for, and in need of, innovation. a

dynamic contextual representation for enterprise

model applications could fill this void and improve

enterprise modelling substantially.

The demand for a business process visualization

that simplifies and individualizes process models is

large, so is the potential for further research in this

field. Contextual representations of

enterprise/process models could be a feasible and

practical way of dealing with the demands of all

stakeholders. The texts introduced here represent the

challenges and expectations researchers face today.

out of them, we can deduct these five core

requirements for our contextual representation for

enterprise model applications (CREMA):

 Effectivity

 Standardization

 Transparency

 Holistic scope

 Complexity reduction

3 RELATED WORK

after reviewing stakeholder demands and needs, we

can identify the increase of effectivity,

standardization, transparency, and the holistic nature

of an enterprise model as main demands, together

with a reduction of complexity. We will now

examine attempts to deal with these demands, while

focusing particularly on context-based approaches.

in “Enabling a User-Friendly Visualization of

Business Process Models” (Hipp et al., 2014),

researchers are trying to find useful forms of

enterprise model visualizations. the different types

introduced in the paper do not differ significantly

from more traditional process model visualizations,

hence their use is limited concerning reducing

complexity. Furthermore, the presented

visualizations seem to be designed for smaller

models; this further reduces their potential to tackle

complexity. The paper also includes a study. This

study is too small in scope and the participants are

too homogenous to make a useful contribution.

Different representations are needed when an entire

enterprise model needs to be visualized, yet they

have to offer real potential benefits.

“ISEAsy“ (Santorum et al., 2014) combines

video games and social media in order to use

visualized processes and business-structures in a

dynamic and accessible way. Though it attempts to

present a holistic view of an enterprise, the actual

visualization doesn’t reduce the complexity as

desired. The introduction of an internal social

network will more than likely also reduce efficiency

within a company. “ISEAsy” does allow for

individual perspectives and has potential to provide

contextual information, yet its visualization would

have to be significantly updated, as it is quite

conservative at this point in time.

Figure 1: ISEAsy Interface (Santorum et al., 2014).

The „PAIs“-Method (Kolb and Reichert, 2013)

tries to solve the problem of complex process

models by utilizing different, personalized views.

The different perspectives are conceptualized as

different levels of aggregation. These aggregations

of process-elements are essential because of the

aforementioned growing complexity of current

business models. The automatic updates which

Reichert and Kolb designed in their paper offer a

large potential for business process modelling. This

aggregation, and its automatization, could

ContextualRepresentationsforEnterpriseModelApplication(C.R.E.M.A.)

485

significantly reduce complexity while also helping

to standardize processes, yet it would not necessarily

provide a more holistic representation of the

company.

Reichert proposes a different tool in another

paper, where he uses Google Earth as a template to

create three-dimensional, contextual and user-

friendly business process visualizations (Hipp et al.,

2012). The implementation of the third dimension

into enterprise models is still in its infancy and there

is a lot of potential for further research. The Google

earth approach is a first step in this direction. This

approach looks at the same process with a varying

degree of details, yet the perspective essentially

remains the same. The method of zooming as well as

panning seems like a natural fit for a comprehensive

enterprise model, yet to estimate the actual

applicability, a practical application of the concept

would be useful.

Figure 2: Google-Earth as business-model Visualization

(Hipp et al., 2012).

The fast delivery of appropriate information to

knowledge workers is an endeavour similar to the

struggle to tackle the growing complexities of

enterprise models: context-awareness is the key to

both. the niPRO framework attempts to deliver

timely information through a context analysis (Hipp

et al., 2013). the approach itself seems to be feasible

for mid-sized models at best though, since it

demands a lot of parameter setting, which would be

unpractical in a larger, more complex model.

The study of the semantics of enterprise and

process models lays down another foundation for

our own work, since it deals with the “language“

aspect of enterprise and process visualization in

depth (La Rosa et al., 2011). the diversity of

possibilities when it comes to visualizations is

important since a different context can call for a

different form of visualization. These different

visualizations can be linked within one model,

which is essential for a holistic view on an

enterprise. Mendling has worked on more studies

which look into other aspects of business and

process models. His work on the careful definition

of process categories (Malinova et al., 2014), as well

as the importance of hierarchical, modular models

(Reijer and Mendling, 2008), are essential for our

own research. The first one calls for particular

attention in the crafting of categories and contexts,

whereas the second emphasizes the need for

abstraction within visualizations.

In “Business Process Modeling: a Multi-

perspective Approach Integrating Variability” five

process views are connected with a process

contextualization (Saidani and Nurcan, 2014). the

five perspectives (intentional, organizational,

functional, non-organizational, non-functional) offer

an incentive to think about the definition of

perspectives within a process-model framework,

since the universal applicability of these

perspectives can be doubted. the inclusion of those

perspectives within a contextualized framework is

important for our own research, as different

stakeholders demand different perspectives.

Additionally, different perspectives that stakeholders

can choose from, would allow for more

standardization.

Under the umbrella of the IsyProM-project

(Jochem et al., 2012), the goal was established to

find a way to display context sensitive abstractions

and perspectives which are based on one core model.

This core model can be individually configured

depending on role, assignment, phase or aspect

towards a user. These contexts can be recognized

automatically and applied to the user’s interface.

Munkhbadarch-Dietrich developed a context

sensitive client for MO

GO which allows for

individual configuration of contexts

(Munkhbadarch-Dietrich, 2012). This client changes

the elements that are visible, but it does not go

further in establishing real individual, context-based

perspectives.

Figure 3: Process visualization without and with activated

action-context (Munkhbadarch-Dietrich, 2012).

Parallel to Munkhbadarch-Dietrich, Gering

(Gering 2012) developed a contextual view on the

Process Assistant. It enables the creation of

MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment

486

individualized perspectives of the Process Assistant,

giving the user a more limited, more understandable

perspective. This is limited to the strictly textual

framework of the process assistant though. Gering

himself criticizes the missing link between the

Process Assistant and MO

GO. Both approaches,

Munkhbadarch-Dietrich’s and Gering’s, exist

independently from one another.

As this selection of texts has shown, there is still

a need for an approach that allows a representation

of an enterprise model, which addresses the most

important stakeholder demands equally.

4 CONTEXTUAL

REPRESENTATIONS OF

ENTERPRISE MODELS

Current methods have seen success in creating

perspectives which use certain filter criteria to

simplify visualizations by aggregating them or by

removing or adding certain elements, yet they

almost entirely remain in the same visualization

style which is graph based. The number of elements

and the complexity of contemporary models demand

other forms of representation. These forms need to

cater to the individual needs and challenges of

stakeholders and have to provide optimal support

while also delivering information in a usable way.

Our Approach is to derive several contextual

representations for the specific stakeholders to

facilitate their specific tasks in the best way.

Therefore we developed a framework for contextual

model representation.

4.1 Framework for Contextual

Enterprise Model Representations

Figure 4: Framework for contextual enterprise model

representations.

The framework for contextual enterprise model

representations comprises of three main elements

(see Figure 4).

The first element is the Context Engine which

identifies all relevant model elements (information

objects). to identify all relevant elements the context

engine uses either a stakeholder (for instance a role)

and/or a given task as input parameters. Based on

this information, the existing enterprise model is

analyzed and all relevant model objects were

identified. This results in a partial model, where not

all single objects need to have a relationship. a

schematic example for that is given in Figure 5.

Figure 5: Schematic example for contextual filtering.

For the identification of relevant model objects

we use different methodologies. One is the tagging

of model elements with a common set of keywords.

Another one is mapping algorithm which connects

typical tasks with their related enterprise model

object types. for instance if a stakeholder is looking

for a specific quality related document the algorithm

automatically maps that task to the documents class,

which includes all quality documents. Furthermore

we are using semantic networks to identify further

related information (model objects) which will be

included in the representation.

The second element is the Template Database.

This database currently includes 8 different

visualization templates for representing enterprise

models:

 Textual

 Process Graphic (currently IEM notation)

 Tree-based

 Organization chart

 Graphical (without a specific notation)

 Network diagram

 Gant chart

 Swim lanes

The appropriate representation is preselected by

the context engine based on its input parameters.

The last element within the framework is the

Transformation Engine. Its uses the set of relevant

model objects which is delivered by the context

engine and builds the contextual representations.

The context engine already preselected possible

suitable visualizations according to the given

stakeholder and/or task, but the final decision is

ContextualRepresentationsforEnterpriseModelApplication(C.R.E.M.A.)

487

made by the transformation engine. This is due to

the aspect that only the transformation engine can

validate and verify if a visualization template can be

applied for the given set of relevant model objects.

4.2 Application

A first application of this framework is implemented

in the Process Assistant (PA), a web-based system

for representation and analysis of enterprise models

build with the Integrated Enterprise Modelling

(IEM) methodology. Currently the PA supports

already some of these visualization templates like

the textual, tree-based and organization chart and

process graphic (IEM notation).

Figure 6: Application of the Framework for Contextual

Representations within the Process Assistant (example

shows graphical, process graphic, tree-based and textual

representations).

5 CONCLUSION AND FUTURE

WORK

The continuously increasing size and the resulting

complexity of enterprise models require new

methods and technologies for the management and

application of said models.

This paper presents a framework for contextual

enterprise model representations in order to provide

different visualizations of the model content. These

representations are derived from the enterprise

model to meet the requirements of different

stakeholders and their specific tasks. Therefore we

use different visualization methods and techniques

to provide all relevant information in the most an

optimal and practical way. This fosters the wide

application of large and complex enterprise models

for different business activities.

For future work we are currently researching

which visualization methods and techniques work

best regarding usability, information content and

mobile applications. Therefore we need to analyze

further input factors which are influencing the

choice of representations and/or the identification of

model elements. Possible new input factors could be

the user’s device, his skills/abilities and personal

characteristics. By allowing the user to select his or

her own emphasis on certain aspects (for example:

certain materials in the overall production or

focusing on cost-intensive areas), we can satisfy the

stakeholders need more individually. We also plan

to implement a mechanism which helps to easily

change the level of detail (e.g. as is possible with

Google Maps). Furthermore we like to include an

easy mechanism to allow the user to create his own

personalized representations.

REFERENCES

BARC, 2013. Steigende Komplexität als

Hauptherausforderung von Performance-

Management-Anwendungen. URL: http://www.find

software.de/Fachinfo/Steigende-Komplexitaet-als-

Hauptherausforderung-von-Performance-

Management-Anwendungen.html [Stand 2014-07-31].

BearingPoint, 2012. Business Process Management-Studie

2012: Stärkung der pro zessorientierung im

Unternehmen durch nachhaltige Optimierung der

Prozess- und IT-Landschaft. URL: http://www.bearing

point.com/de-de/7-5990/business-process-

management-studie-2012-white-paper/.

Bobrik, R., 2008. Konfigurierbare Visualisierung

komplexer Prozessmodelle. 1. Aufl. München: Dr.

Hut. (Informatik).

Gering, P., 2012. Analyse zur Anwendung von

kontextuellen Sichten eines modellbasierten

Prozessassistenten im Rahmen der Unternehmens-

modellierung anhand eines Anwendungsfalls aus der

Luftfahrtindustrie. Studienarbeit. Technische

Universität Berlin.

Harmon, P., Wolf, C., 2014. The State of Business Process

Managment 2014. URL: http://www.bptrends.com/

bpt/wp-content/uploads/BPTrends-State-of-BPM-

Survey-Report.pdf [Stand 2014-07-31].

Hipp, M., Mutschler, B., Reichert, M., 2011. on the

Context-aware, Personalized Delivery of Process

Information: Viewpoints, Problems, and

Requirements. In 6th Int’l Conf. on Availability,

Reliability and Security (ARES’11), Industrial Track,

390–397.

Hipp, M., Mutschler, B., Reichert, M., 2012. Navigating in

Process Model Collections: a New Approach Inspired

by Google Earth. in Aalst, Wil, u.a. (Hg.): Business

Process Management Workshops. Berlin, Heidelberg:

Springer Berlin Heidelberg. (Lecture Notes in

Business Information Processing), 87–98.

Hipp, M. et.al., 2013. a Framework for the Intelligent

Delivery and User-Adequate Visualization of Process

Information. In 28th Symposium on Applied

MODELSWARD2015-3rdInternationalConferenceonModel-DrivenEngineeringandSoftwareDevelopment

488

Computing (SAC’13): ACM Press, 1383–1390.

Hipp, M., et.al., 2014. Enabling a User-Friendly

Visualization of Business Process Models. in. 3rd Int’l

Workshop on Theory and Applications of Process

Visualization (TaProViz’14), BPM 2014 Workshops.

Houy, C., Fettke, P., Loos, P., 2012. Understanding

Understandability of Conceptual Models – What Are

We Actually Talking about?. in Hutchison, David, u.a.

(Hg.): Conceptual Modeling. Berlin, Heidelberg:

Springer Berlin Heidelberg. (Lecture Notes in

Computer Science), 64–77.

Jochem, R., Knothe, T., Wintrich, N., 2012. Enforcing

front-loading in engineering processes through

product-process integration. in Enterprise

Interoperability V. Guildford, Surrey: Springer

London, 71–80.

Kolb, J., Reichert, M., 2013. Data flow abstractions and

adaptations through updatable process views. in Shin,

Sung Y. & Maldonado, José C. (Hg.): the 28th Annual

ACM Symposium, 1447.

La Rosa, M., et. al., 2011. Managing Process Model

Complexity via Concrete Syntax Modifications. in

Industrial Informatics, IEEE Transactions on 7(2),

255–265.

Malinova, M., Dijkman, R., Mendling, J., 2014. Automatic

Extraction of Process Categories from Process Model

Collections. in Lohmann, Niels, Song, Minseok &

Wohed, Petia (Hg.): Business Process Management

Workshops. Cham: Springer International Publishing.

(Lecture Notes in Business Information Processing),

430–441.

Mendling, J., Reijers, H., Cardoso, J., 2007. What Makes

Process Models Understandable? in Alonso, Gustavo,

Dadam, Peter & Rosemann, Michael (Hg.): Business

Process Management. Berlin, Heidelberg: Springer

Berlin Heidelberg. (Lecture Notes in Computer

Science), 48–63.

Munkhbadarch-Dietrich, M., 2012. Entwicklung eines

Clients zur Bereitstellung kontextueller Sichten von

Unternehmensmodellen anhand eines

Anwendungsfalls im Maschinenbau. Diplomarbeit.

Technische Universität Berlin.

Reijers, H., Mendling, J., 2008. Modularity in Process

Models: Review and Effects. in Dumas, Marlon,

Reichert, Manfred & Shan, Ming-Chien (Hg.):

Business Process Management. Berlin, Heidelberg:

Springer Berlin Heidelberg. (Lecture Notes in

Computer Science), 20–35.

Saidani, O., Nurcan, S., 2014. Business Process Modeling:

a Multi-perspective Approach Integrating Variability.

in van der Aalst, Wil, u.a. (Hg.): Enterprise, Business-

Process and Information Systems Modeling. Berlin,

Heidelberg: Springer Berlin Heidelberg. (Lecture

Notes in Business Information Processing), 169–183.

Santorum, M., Front, a, Rieu, D., 2014. ISEAsy: a Social

Business Process Management Platform. in Lohmann,

Niels, Song, Minseok & Wohed, Petia (Hg.): Business

Process Management Workshops. Cham: Springer

International Publishing. (Lecture Notes in Business

Information Processing), 125–137.

Schwermer, M., 1998. Modellierungsvorgehen zur

Planung von Geschäftsprozessen. Berlin: IPK

(Berichte aus dem Produktionstechnischen Zentrum

Berlin).

Vernadat, F.B., 2002. Enterprise modeling and integration

(EMI): current status and research perspectives. in

Annual Review in Control 26: 15-25.

ContextualRepresentationsforEnterpriseModelApplication(C.R.E.M.A.)

489