Analysis and Evaluation of Business Process Modeling Adoption in
Collaborative Networks
Hodjat Soleimani Malekan and Hamideh Afsarmanesh
Informatics Institute, University of Amsterdam, Post bus 94323, 1090 GH Amsterdam, The Netherlands
Keywords: Business Processes, Collaborative Network, Business Process Modeling Languages, Virtual Organizations
(VO), VO Breeding Environments, Service-Oriented Architecture.
Abstract: Besides, knowledge and information enterprises can share Business Processes (BPs) within Collaborative
Networks (CNs). Each enterprise has a set of BPs that it can perform, and through developing integrated
BPs in the CN they deploy their capacities and capabilities. Selecting and adopting the appropriate BP
modelling languages (BPML) for the purpose of formalizing BPs are challenging, because of the variety of
existing methods, tools, and standards with different strengths and weaknesses. In surveys published so far
on BP modeling mostly, a set of general features of the main BP languages and standards are compared.
However, they have not paid attention to the level of different categories of BPMLs. Furthermore, there are
no surveys analysing and evaluating the prerequisites to fulfil CN’s requirements. This paper first proposes
a set of categories for the main BP languages and standards. Then a novel BP evaluation approach, in CN
context is introduced. Finally, different categories are discussed and analysed by addressing their suitability
to support CNs.
1 INTRODUCTION
Adopting Business Process Modeling Languages
(BPMLs), including the introduced standards, tools,
and techniques, have greatly influenced enterprises
toward capturing opportunities, reducing costs, and
increasing productivity.
The BP technologies themselves however have
also been affected by high demand of market, as
well as the step-wise maturity of Business Process
Management (BPM) theories. This has caused rapid
changes in the last decade developed BPML tools
and standards, while creating challenges for the BP
modeling selection and adoption in networked
enterprise. (Camarinha-Matos and Afsarmanesh,
2008).
Most of these aproaches are founded on Service
Oriented Architecture (SOA), apply the formalized
BPs, performed at every enterprise, to facilitate
service interoperation and enterprise collaboration
(Papazoglou and Heuvel, 2006). formalized BP are
therefore important for effective cooperation among
different enterprises within the Collaborative
Networks (CNs), and without formalized
representation of their BPs, enterprises cannot
effectively share their competencies and capabilities.
The BPMLs differ from each other in their
modeling approaches for design, analyzing, and
enacting of BPs. Focusing on the purpose of
supporting enterprises, with their collaboration
within the CNs, the selection and adoption of a
suitable BPML is critical, while challenging.
Published surveys on BPMLs e.g. Roser and Bauer
(2005), LU and Sadiq (2007), also Ko, S.Lee and
E.Lee (2009) have already tackled the comparison
between a certain features of the main BPMLs tools
and standards.
Most contemporary surveys focus on comparing
a set of two or more BPML standards and tools. So,
there is a lack of emphsis on comparing different
categories of BPMLs, to which these standards or
tools may belong. For instance the distinct features
aimed by their design, such as to evaluate and
emphasize their graphical, ontological, executional,
etc. aspects of the BP modeling, is not assessed for
this purpose.
Moreover, demonestrating a set of categories for
BPMLs classification, in order to perform analysis
and evaluation of BPMLs for their adoption in
support of CNs, a novel analysis method is
introduced to manifest CN’s characteristics and to
assess different BPML categories against them.
23
Soleimani Malekan H. and Afsarmanesh H.
Analysis and Evaluation of Business Process Modeling Adoption in Collaborative Networks.
DOI: 10.5220/0004773700230032
In Proceedings of the Third International Symposium on Business Modeling and Software Design (BMSD 2013), pages 23-32
ISBN: 978-989-8565-56-3
Copyright
c
 2013 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved

Hence, we first review the main concepts of CNs
and BPs, identifying the role of formalized BPs in
CN (in Section 2). Then, based on a systematic
reviewing approach, and considering the existing
categorizations of the main BPMLs, we introduce
our BPML categorization (in Section 3). In Section
4, founded on collaboration purposes, we specify a
number of most relevant criteria for comparing the
introduced BPML categories, and analyzing them
for the aim of supporting enterprise collaborations.
Finally, our analysis and evaluation approach is
discussed (in Section 5), and our conclusions are
presented (in Section 6).
2 ROLE OF BPs IN CNs
Within the collaborative-networked environment the
enterprises have the opportunities to share their
resources through collaboration, including
knowledge, information. This can be best achieved,
by means of formalized BPs (Camarinha-Matos and
Afsarmanesh, 2008). Collaborative BP integration is
aimed by enterprises to accomplish value-added
business services, beyond the capabilities of their
individual organizations.
Besides integration, in most approaches for
instance presented by Papazoglou and Heuvel
(2006), BPs constitute the building blocks for
establishment of SOA, through BPs implementation
as web services. In this section, after reviewing
related definitions for CNs and BPs, we present an
analysis of the BPMLs from the CN requirements
point of view.
2.1 Principal Definitions
A general definition of Collaborative Network is
presented by Camarinha-Matos and Afsarmanesh
(2008) as: “an alliance constituting a variety of
entities that are autonomous, geographically
distributed, and heterogeneous in terms of their
operating environment, decision making, culture and
social capital, that cooperate/collaborate to better
achieve common/compatible goals, and their
interactions are supported by the computer
networks.”
The two main forms of CNs are the: Virtual
Organization (VO) and VO Breeding Environment
(VBE). In a VO, partners choose co-working and
sharing their BPs and other resources to accomplish
their common goals. The motivations for this
coalition are commonly formed around specific
market targets or innovation purposes. VBEs, which
establish long-term alliances of organizations,
capture and save BPs of partners in their directories.
The VO broker, who seizes the opportunity and
chooses the participants for the VO in the VBE
context, considers selecting and integrating BPs of
different organizations to shape new VOs
responding to achievable opportunities.
(Afsarmanesh et al. 2011).
Related to our research, a set of standard
definitions for the BP notions exist that is provided
by Workflow Management Coalition (WFMC,
1999) and is addressed below.
A typical definition of BP is: a series of one or
more linked procedures or activities, which
collectively realize a business objective or policy-
related goal. Workflow Management System
(WFMS) can automate and control the execution of
the BPs. The notion of BPM comprises concepts,
methods, and techniques to support organizational
aspect of processes, which are needed for the design,
administration, configuration, enactment, and
analysis of BPs (Weske, 2007). It also covers the
“diagnosis” aspect of the BPs further to the WFMS
lifecycle (Van der Aalst, 2003).
Havey (2009) outlines the focuses of BP
modeling on design and execution aspects of the
BPs. BP Modeling aims at representing an abstract
but meaningful demonstration of the real business
domains. This goal is achieved through provision of
appropriate syntax and semantics in BPMLs, to meet
the BP’s requirements. (Lu and Sadiq, 2007).
2.2 Chronological View of BPMLs in
Support of Collaboration
Here we address the evolution of BPMLs from
collaboration point of view. In the 80s, the necessity
of process-awareness was recognized, beyond the
level which was required for development of
Management Information Systems (MIS).
Furthermore, besides understanding the flow of
operations in MIS, organizations and business
domain experts needed to also understand the
information aspects of the BPs in MIS (Delvin and
Murphy, 1988).
The WFMSs, which initially were intended to
facilitate automatic transformation of electronic
documents, was then introduced as the new tools to
enable business analysts in designing and expressing
BPs, at the beginning of 90s. For the purpose of
depicting information exchange among systems, the
behavioural concepts (i.e. the sequence and merge)
were then used in BP modeling (Georgakopoulos,
1995).
Third International Symposium on Business Modeling and Software Design
24

Afterward in the 90s, applying the Business
Process Re-engineering as well as embedding the
best business practices in the market, vendors were
able to integrate and aligned separate software
modules, under the so-called Enterprise Resource
Planning (ERP) systems. To support ERPs, the
BPMLs have focused on dynamic aspects of the
BPs. Nevertheless the interactions between the
designed modules were not so easy to achieve within
the ERPs (Van der Aalst, 2009).
Responding to the proliferation needs of the
integrating legacy systems into customized
applications and ERP modules, the Enterprise
Application Integration (EAI) (Lee and Siau, 2003)
have tried to remedy the problem of inefficient BPs’
integration. So, interaction-enabling entities (e.g.
messages) gained significance. This level of
collaboration provided an infrastructure for
cooperation of enterprises through resource sharing,
while preserving their heterogeneity.
The more maturity in deployment of XML, in the
late 90s, resulted in better integration of
applications, and changed the co-working intensity
of enterprises to an advanced level, called business
to business (B2B) (Havey, 2009).
Coordinating the BPs adopted by companies,
concluded in integrating autonomous and
independent applications, via loosely coupled
mechanism of SOA (Zdun et al., 2006). SOA
approach tries to establish orchestration and
choreography of web services, to achieve their
successful cooperation.
Nowadays, BP related topics e.g. the BP mining
(Van der Aalst and Dustar, 2012) and diagnosis
approaches that address BP monitoring and their
continuous improvement constitute promising
research lines.
3 CATEGORIZATION OF BPMLs
Aiming to cover various BP modeling tools and
standards, which are introduced in the main related
publications, we focus on a specific set of attributes
and specifications of the BPMLs for their
categorization. Our categories basically focus on
recognizing the BPML’s capabilities as well as the
suitability features in each category, in support of
criteria for collaboration. Therefore, we first study
related scientific DB and conferences, then classify
the existing categorization publications into two
classes of: “General Review”, and “Particular
Evaluation”.
In this section, first we review the results
presented in published surveys, from the point of
view of our two classes addressed above, and further
classify a set of minimal relevant BPML
categorization approaches. Finally, we introduce our
more detailed categorization.
3.1 BPML Categorization Review
As mentioned before, we divide the contemporary
reviews of BPMLs into two main classes of
“General Reviews” and “Particular Evaluation”.
“General Reviews” are mostly focused on general
uses, and on encompassing the main specifications
of the BPML categories. For instance the work of
Havey (2009) that focuses on presenting good BP
Modeling Architecture, where it first addresses
aspects of BP modeling applications (i.e. design,
run, monitor, etc.), and then introduces the four
categories of BPMLs, including: notation languages
(e.g. BPMN), execution languages (e.g. BPEL),
choreography languages (e.g. WS-CDL), and
process administration languages (e.g. BPQI). Also,
classification presented by Ko et al. (2009) and Mili
et al. (2010) are instances of this category.
But, publications in the “Particular Evaluations”
class focus on BPML categorization for specific
purposes. The works of Roser and Bauer (2005), Lu
and Sadiq (2007), and De Nicola et al. (2007) are
instances in this category. For example, in (De
Nicola et al., 2007) the categories are introduced
around the subject of “introducing an ontological
approach for BP modeling”, including Descriptive
(e.g. BPMN), Procedural (e.g. XPDL), Formal (e.g.
PSL), and Ontology-based (e.g. OWL-s).
3.2 Introduced BPML Categories
Using the “general review” and “particular
evaluation” criteria as the base, we introduce a more
comprehensive framework including six classes:
“graphical”, “formal”, “executional, “ontological”,
“inter-operational”, and “monitorial”, that together
capture all kinds of addressed BPMLs.
The main characteristic of each of these six
categories, and their main representative example
BPMLs are briefly (due to space limitation)
described in the following subsections. Also, a set of
popular BP languages and standards are named
below as the example of each category.
Although it is possible for a BPML to be
categorized in more than one category, but here we
have placed each BPML in its most representative
category only. They could be adopted and utilized
by CN members based on their category’s
Analysis and Evaluation of Business Process Modeling Adoption in Collaborative Networks
25

Figure 1: The BPML categorization meta-process method.
characteristics.
The meta-process adopted by our categorization
method and how we reached the six specific classes
is briefly depicted in figure1.
3.2.1 Graphical BP Languages
Rooted in graphical picturesque format, this classical
generation has appeared. BP modeling languages in
this category mostly emphasize illustrating the
system behaviour and its abstraction. These
languages are not typically formal. (e.g. IDEF, EPC,
UML 2.0, BPMN).
3.2.2 Formal BP Languages
Formalization in this category is founded upon
mathematical principles. Although, adoption of
graphical symbols is possible in some of these
languages, but difficulties in user’s understanding
hold them mostly at theoretical and mainly academic
utilizations. (e.g. Petri-Net, Pi-calculus, PSL, Reo).
3.2.3 Executional BP Languages
The idea of automatic execution of BPs by software
engines, support the formation of this category. The
XML structure plays an important role in
deployment of this category, and clarifies BPs by
their computerized semantics. Besides, the
popularity of BP modeling and service invocation in
industries are other important issues in the category.
(e.g. BPEL, WS-CDL, XPDL, YAWL).
3.2.4 Ontological BP Languages
Likewise the ontology approach, which studies the
things that exist and tries to describe them, this
category addresses semantic capture and tries to
constitute the base for an increasing number of BP
modeling languages, through proposing different
meta-models. The ontological layer in these
languages clarifies the roles, entities, and
interactions. This category has also the advantages
of using XML formats. (OWL-s, WSMO, BPDM).
3.2.5 Interoperational BP Languages
Rooted in business-to-business interaction, this
category focuses on modeling public sharable
processes of partners, among many business
partners. To accomplish this key concern in inter-
operational category, XML standards are elaborated
as the main enablers. (e.g. RossettaNet, eb-
XML/BPSS).
3.2.6 Monitorial BP Language
As we discussed previously (in section 2), modern
business process modeling trends to address the
diagnosis iteration of the BP Lifecycle. Focusing on
the Business Activity Monitoring (BAM) point of
view, the emphasis is on monitoring and resolving
the deadlocks or problems in flow of BPs.
Furthermore, extract and unambiguous approach for
recognizing BP modeling based on a dynamic
logging of process behaviour, the so-called process
mining is still promising (van der Aalst and Dustar,
2012), (e.g. BPRI and BPQI).
BPML Categorization Meta-Process
Categorize
BPMLs
Represents Graphical
Category
Uses graphical
representation
for BP flow
Represents Formal
category
Uses
Mathematical
Foundations
Uses textual
block-orientation
Uses a meta-model
Ontology for
BP definition
Requires
XML
Formats
Represents
Ontological
category
Represent
Monitorial Category
Facilitates
monitoring
and Mining of BPs
Represent
Interoperational
Category
Extends
Interactions
of
Collaborative
processes
Third International Symposium on Business Modeling and Software Design
26

4 SUPPORTING CN - BPML
EVALUATION
The evaluation framework should be concise and
descriptive. Having emphasis on categories and not
every BPML, multi-aspect evaluation of a
phenomenon requires a methodology, to support
maximal coverage of the target area. For the purpose
of appraising BPMLs in supporting CNs
requirements, we should consider both the BPs and
the CNs aspects simultaneously. Therefore, our
designed evaluation methods as well as our
evaluation process are discussed in following sub-
sections, respectively.
4.1 Proposed Evaluation Method
Several BP modeling goal-settings have been
introduced based on different approaches. For
instance a set of five generic software process
modeling objectives have been specified in (Curtis et
al., 1992) as follow: “to support process
improvement”, “to facilitate human understanding
and communication”, “having automated guidance
in performing process”, “to support process
management”, and “to automate execution support”.
Also, for the non-functional BP modeling
requirement (Chung and Do Prado, 2009) has
presented a series of objectives (e.g. the support for
discovering of dependencies of processes, the
support for changing management, etc.). These
context-aware objectives still hold today. In our
point of view for supporting more effective BP
collaboration in a CN, we can further add, “to
support enterprise collaboration” into this context.
Rooted in the debate in Section 2, our primary
aim is to focus on supporting collaboration through
formalized BPs and evaluating BPML categories for
this purpose. Therefore, we first follow a goal-based
approach (also known as the objective-based
approach) explained in (Goldkuhl and Lagsten,
2012) to extract the collaborative intention issues
within CN’s concept.
Our goal-based approach has focused on a
number of qualitative criteria and indicators, related
to set goals, systematically. As the evaluation
method, we adopt Critical Success Factors (CSFs)
method. CSF is a classical flexible method to
maximize goal achievement, through selecting,
working, and monitoring a few certain factors,
which are vital for success. So, we follow the
requirements of achieving established objectives, by
running a Critical Success Factors Analysis
described and explained in (OASIS, 2008) and
partially in (Trkman, 2010) and (Sudhakar, 2012).
After CSF identification, a set of requirement
indicators for monitoring them is provided by CFA.
The CFA constitutes following elements:
• Objectives: Those are directed by customers and
are hard to measure.
• CSFs: including between three to six sub-goals,
which without their direct support, achieving
goals are unattainable.
• Requirements Indicators: represented key
performance indicators, which are measurable
and directly support CSFs.
• CFA Diagram: for better illustrating the
measurable context, CFA diagram is used.
For supporting characteristics of CNs to achieve
their goals and to better describe the particularities
of the CN context, especially for VOs and VBEs, we
use the Reference model for Collaborative
Networks” (ARCON). The ARCON model explains
aspects, approaches and elements of the CN’s
environment (Camarinha-Matos and Afsarmanesh,
2008).
Eventually, we hold a CFA study to find out CN-
compliance CSFs, and the vital requirements
indicators for achieving our goal. These issues are
provided based on systematic technical reviews and
experts opinions. We then discuss, the BPML’s
categories versus the recognized requirements
indicators, and represent the conclusions.
4.2 Evaluation Process
Regarding the (ARCON) model introduced by
Camarinha-Matos and Afsarmanesh (2008), and our
discussion in section 2, VO/VBE need to manipulate
formalized BPs.
Also, regarding CN’s definition (in 2.1) the
following aspects indicate the main constitutional
objective themes in the CN discipline:
• Goal-orientation [focusing on goals through
business interactions]
• Infrastructure for Commonality [supporting
co-working and coordination toward goals]
• Node Heterogeneity [non-uniformity in
different properties, i.e. operational processes]
• Network enabling [support by computer
networks]
The four above-mentioned objective themes are the
main CN realization’s objectives, extracted from its
standard definition. To attain these objectives,
defining and aligning a set of CSFs are inevitable.
The supporting CSFs for CN are as follows: first,
to enable successful collaboration, BP modeling tool
Analysis and Evaluation of Business Process Modeling Adoption in Collaborative Networks
27

should
p
artners
use” is
interope
r
b
ehavio
u
enactme
n
achieve
m
docume
n
coverag
e
are illus
t
intersec
t
minimal
constitu
t
Tabl
e
CSFs
Co
m
E
Exp
r
A
The
l
of gene
r
context
categori
e
the liter
a
Unders
t
capture
b
user ca
n
develop
Express
the pr
o
resource
(Kiepus
z
Flexibil
i
modelin
g
instance
s
(Lu and
does no
t
of BP m
o
Availab
i
b
usiness
formats,
They ar
e
the orga
n
Enacta
b
p
rovide en
o
(BP Analyst
s
another issu
e
r
ation throu
g
ur
”, is an
o
n
t of BPs
m
ent of goal
n
ts and stand
a
e
of C
N
’s obj
e
t
rated in tabl
e
t
ion of row
s
coverage
b
t
ional objecti
v
e
1: Intersectio
n
C
N
Objecti
v
m
prehensibility
E
ase of Use
r
essiveness for
behaviour
A
ccessibility
l
ast step of o
u
r
ic required
to appraise
e
s for CN’s.
T
a
ture and the
s
t
andabilit
y
: i
b
y which un
d
n
interpret an
the BP mode
l
abilit
y
: expl
a
o
cess mode
l
s, flow struc
t
z
ewski, 2002)
i
t
y
: is define
d
g
the modific
a
s
, based on i
n
Sadiq, 2007)
.
t
necessarily
o
dels (Schon
e
i
lit
y
: compri
s
process m
o
and standar
d
e
ready-to-us
n
izations (Mi
l
b
ilit
y
: is defi
n
o
ugh “comp
r
s
, IT experts,
e
e
, which su
p
g
h CNs. “e
x
o
ther challe
n
in CNs. F
in CNs, “ac
c
a
rds has to b
e
e
ctive and th
e
e
1. The “C”
s
and colum
n
b
etween our
v
e elements.
n
of CN’s obje
c
N
’s
v
es
Goal-Orientation
Commonality
C
C
C
ur
evaluation,
indicators fr
o
the suitabi
l
T
hey have
b
e
e
s
tandar
d
s (IS
O
s the ease of
d
er specified
c
instance, m
o
l
(Mendling e
t
a
ins the capa
b
l
’s attribute
s
t
ures, in an
u
.
d
as the ease
w
a
tions are po
s
n
complete le
v
.
So,
p
artial e
imply compl
e
e
nberg et al.,
2
s
es amount an
o
deling docu
m
d
s are accessi
b
e for desired
l
anivic, 2008)
n
ed as the am
b
r
ehensibility”
e
tc.). The “ea
s
p
ports conve
n
x
pressiveness
n
ging issue
or cos
t
-effe
c
c
essibility” o
f
e
considered.
e
introduced
C
in the box a
t
n
s represent
CSF and
C
tives and CSF
s
Commonality
Heterogeneity
Network enabling
C
C
C
C
C
is finding a s
e
o
m BP mod
e
l
ity of BP
M
e
n extracted
f
O
, 2010E):
i
nterpretation
c
i
r
cumstances
o
del, analyse,
t
al., 2007).
b
ility to repr
e
s
like: co
n
u
nambiguous
w
ith which i
n
s
sible in types
v
el of abstra
c
ffects of cha
n
e
tely replace
m
2
008).
d
degree to
w
m
ents in spe
c
b
le and adopt
a
collaboratio
n
.
b
ition of acqu
i
for
s
e of
n
ient
for
for
c
tive
f
BP
The
C
SFs
t
the
the
C
N’s
s
.
Network
enabling
C
e
ries
e
ling
M
L’s
f
rom
n
and
, the
and
e
sen
t
n
trol,
way
n
BP
s
and
c
tion
n
ges,
m
ent
w
hich
cific
a
ble.
n
by
i
ring
ca
p
dir
e
inf
o
int
e
CS
F
the
(w
h
arr
o
eac
sta
n
5
Ou
r
des
of
B
are
foc
u
ori
g
a n
u
cat
e
5.
1
Av
a
im
p
ser
v
20
1
ha
p
“in
t
(M
i
ass
u
do
c
20
0
p
ability to co
m
e
ctly and w
i
o
rmation (Ru
s
Figure 2: I
n
Figure2 re
p
e
ractions of d
i
F
s and moni
t
types of effe
h
ich is repre
s
o
w). In the n
e
h
of BPM
L
n
dard criteria.
DISCUS
r
evaluatio
n
criptive eval
u
B
PML categ
o
introduced
a
u
s on novel
g
ins – instea
d
u
mber of cho
i
e
gories for C
N
1
Availa
b
a
ilability has
p
lies, ratio of
t
v
ice accordin
g
1
0E). Unavai
l
p
pens when
t
ervals”, a
n
i
lanivic et a
l
u
me availab
i
c
uments withi
n
There is an a
n
0
5, by Wolf
m
pletely exe
c
i
thout explo
i
s
sell, 2007).
n
terconnection
s
p
resents CF
A
i
fferen
t
CN’s
t
oring requir
e
ct
b
etween d
i
ented in the
e
xt section w
e
L
s categoriza
t
SION
n
comprises
u
ation. The fi
o
ries. Six co
m
a
nd defined.
BPML cate
g
of the langu
a
i
ces from BP
M
N
s.
b
ilit
y
its roots in r
e
t
he time that
u
g
to prior lev
e
ability of a
m
we don’t
n
d “user-
p
e
r
l
., 2008). F
o
i
lity as the
n
the context
o
n
nual researc
h
and Harmo
c
ute of the
B
i
ting extra t
s
in CFA diagr
a
A
analysis
objectives, s
u
e
ments indic
a
i
fferent entiti
e
figure by th
e
e
address an
d
t
ion against
a two-di
m
i
rs
t
dimensio
n
m
prehensive
c
Please note
g
ories – as t
h
a
ges, so ther
e
M
L’s membe
r
e
liability noti
o
u
se
r
s have re
c
e
l of agreem
e
m
odeling BP
have “stea
d
r
ceived” a
v
o
r our evalu
a
existence o
of CNs.
h
of BPM Ma
r
o
n (2012), t
h
B
P model
ools and
a
m.
as the
u
pporting
a
tors and
e
s in CNs
e
support
d
evaluate
the five
m
ensional
n
consists
c
ategories
that we
h
e aimed
could be
r
s in each
o
n, which
c
eive
d
the
e
nts (ISO,
language
d
y-state”,
v
ailability
a
tion, we
f BPML
r
ket since
h
oroughly
Third International Symposium on Business Modeling and Software Design
28

surveying the BPM trends. From their 2012 report,
they state that the rate of availability for graphical
BPMLs is at the highest level. For example the
BPMN is used by more than 60% of all
organizations. Meanwhile, there is less availability
for ontological BPMLs (e.g. BPDM). Although, the
debate on the timely development of trends is not the
focus of this paper, but decrease in attraction level of
BPEL during recent years is noticeable. Even
interest and availability of UML and EPC slightly
decrease. Also according to that survey the
pervasiveness of the rest of BPML categories (e.g.
interoperational, formal) are the lowest in usage
ranking. So, it is expected that organizations initiate
collaboration in CNs applying graphical BPMLs,
and especially BPMN.
5.2 Enactability
The enactability is an important phase in BPM life
cycle. According to (van der Aalst et al., 2003), after
accomplishing “process design” and “system
configuration” at the third step of the BPM’s
lifecycle “process enactment” is located right before
the “diagnosis” step. The more independent is the
BPML from the technology and vendor executable
environments, the better enactability has in CNs.
Using the formal semantics for more effective
enactment (ter Hofstede et al., 2010) supports – and
does not contradict the increase in understandability
in support of the requirements in CNs. Executional
BPMLs enable the enactments of BPs, for sharing
BP’s information and automatically executing them
through block-based and machine understandable
structures. But, despite their common executional
capabilities, they have their particularities.
Within Executional category, BPEL describes
behaviour of BPs within interaction between process
and its partner, and efficiently supports
orchestration. WS-CDL executional aspect consists
of peer to peer collaboration of partners from a
global point of view for supporting choreography.
But, in this category some of the languages such
as BPEL have restrictive syntax (Recker and
Mendling, 2006), which is a limitation for this
popular language, and some (e.g. YAWL) have
exact executional syntax (ter Hoftstede 2010).
The formal category languages - except
embedded notions like (pi-calculus in WSCDL)
provide graphical enactability interface, e.g. in reo
and Petri net. Ontological BPMLs, because of their
XML supporting structures have convenient level of
enactability.
Executional issues in interoperational BPML
category, where XML enactability is embedded,
have some difficulties e.g. naming and XML
reusability in RosettaNet (Damodaran, 2004) or
deficiencies in event handling during interactions
(Green et al., 2007). Ontological BPML category
focuses on semantic aspects (e.g. OWL-s), and runs
enactment in an abstract level.
5.3 Expressability
The importance of expressability in CNs arises from
the way we wish to express the BPs, so that they can
be shared among partners. This expressive power of
modeling language represents the possibility of
expressing constructs in direct or indirect manner
(Kiepuszewski et al., 2002). These constructs
comprise: control, resources, data, organization,
execution, and behaviour of a business models.
Expressability encompasses the notion of suitability,
which focuses on modeling and implies
conformance of the BPML with for instance 43
workflow patterns introduced in (Russell et al.,
2006). Although, the evaluated domain in that paper
does not focus on BPML categories, but provides a
general inception for comparison of BPML
categories.
As we map BPMLs’ evaluation in Russell, ter
Hofstedeh and van der Aalst, (2006) to our proposed
categories, a number of these patterns e.g.
“discrimination”, “milestone”, “partially join”, etc.
are the kind of patterns which languages and
standards have difficulties in expressing them.
We could state that, commonly, the graphical
BPML category has better compatibility, while in
executional category- except for YAWL- languages
have some deficiencies, for example for supporting
“Arbitrary Cycle”, because of their rigidity in
capturing real-world abstraction.
Based on evaluation of Russell et al., (2006) the
of formal languages category members have good
capability of expressiveness, because of their
mathematical foundation, e.g. Petri-Net; expressive
power (van der Aalst et al., 2003) used in workflow
pattern design, or constraint Automata is used in
Reo. Ontological languages use logical basis for
instance in OWL-s for representing better
expressiveness (W3C-OWL-s, 2004).
5.4 Flexibility
Supporting the dynamicity of CNs, the flexibility
issues in BPMLs for describing BP’s interaction is
necessary. BPMLs focus to sustain their dynamicity
in coping with expected and unexpected changes,
Analysis and Evaluation of Business Process Modeling Adoption in Collaborative Networks
29

through adopting flexibility. In (Schonenberg et al.,
2008) four types of flexibility are presented as:“
design”, ”deviation”, ”underspecification” and
“change”.
The flexibility support, mostly in two first above-
mentioned types, BPMLs rely on their pre-design
notations and are abstract from flexibility concerns.
On the other hands, the block-based (rule-based)
BPMLs could manage the flexibility in higher level
(e.g. deviation or underspesification) (Lu and Sadiq,
2007).
The flexibility in the graphical BPML category,
within different languages and standards is
considered in different ways. In BPMN, by
predicting three types of diagram for collaboration,
and for the concepts of Pool and lane, the
decomposition for changes is possible. The Frame
and Frame Heading techniques in UML 2.0 Activity
Diagram let the elements of the languages to be
defined and described in a modular and flexible
structure. So, “design and deviation” are supported.
Likewise in formal category, mathematical
concepts help to retain model identity; for instance
the structure of Atomic and Complex activities in
PSL, besides graphical representation in Petri-Net
and Reo simplifies the modification flexibilities. So,
“design and deviation” are supported.
Based on XML structures, which usually support
flexibility in design and changes, and even
underspesification, to certain extent (Schonenberg
2008). YAWL, BEPL (inter-relations), and WSCDL
(choreography) support various types of flexibility.
Even RosettaNet PIP techniques, channelizes the
modifications. This benefit supports within block-
based structure. Ontological BPML category
considers flexibility at convenient level, which let
modification to be based on primary definition of
BPs (e.g. process model definition in OWL-s).
5.5 Understandability
The understandability shall facilitate the BP
acquisitions and interactions among CN’s
stakeholders. This notion has been reviewed and
analysed during several works especially verses the
complexity as the other extreme. Generally,
understandability comprises the following two
aspects mentioned in (Mendling et al., 2007):
• Model-related factors, which affect the
understandability, e.g. unambiguity, simplicity.
• Person-related factors, which have close relations
to knowledge and experience of participants
Although, the understandability has been reviewed
several times, and there is a number of guidelines
e.g. the smaller size of the model makes it better for
understanding; or the higher degree of input and
output to one element causes the more complexity of
understandability, etc. But, the ease of
“comprehension of a model”, “presenting without
error”, and “labelling less ambiguous” (Mendling et
al., 2010) constitute main understandability’s
principles in BPMLs.
Generally, the graph-based languages are more
understandable than rule-based ones (Lu and Sadiq,
2007). That is also the reason why they become
more popular at enterprises. However, within
graphical standards, BPMN is more complex for
understanding compared to UML and EPC (Indulska
et al., 2009).
On the other hand, Executional and ontological
BPML categories because of having less cooperation
with human side, their understandability is under
criticism. Also, the interoperational standards (e.g.
the PIP technics knowledge in Rosettanet) are at a
more abstract level of understandability
((Damodaran, 2004), (Green et al., 2007)).
5.6 Comparing Results
Through the discussion, we analysed the adoption of
BPML in regards to the set of requirement
indicators, which represented for evaluating BPML
categories at the second level of our evaluation.
Grounded in our goal-based approach and by using a
CFA method, we identified six requirement criteria
that helped us, to measure the collaboration-aware
adoptability of BPMLs. The result of our extensive
evaluation in previous sections is summarized in
Table 2.
Table 2: Summary of evaluation.
Requirements
indicators in
Support
Of CN
BP
Modeling
Languages
Categories
Understandability Expressibility Enactability Availability Flexibility
Graphical S A A S A
Formal M S A M A
Executional A A S M A
Ontological A S A A A
Interoperational A A S A A
Monitorial N N N N N
As showed in that table, we use four levels of
Comments:
S: Strong support A: Advanced support
M: Moderate support N: Not Addressed
Third International Symposium on Business Modeling and Software Design
30

supports as: Strong, Advanced, Moderate and Not
addressed levels, from CN’s member relative points
of view. Because of analytical theme of paper, we
opt qualitative survey method.
Regarding discussions in our previous sections,
the graphical category has advantages of
understandability and availability. Executional
category is strong in enactability, also flexibility of
BPs, besides the importance of less ambiguity in
modeling real world should not be disregarded,
although lacks of interactive graphical depiction
needed for less technical users is serious criticism
yet.
Due to complexity of their user interaction, the
formal languages are not pervasive, but should be
considered as the supporting layer for soundness for
graphical modeling languages. Ontological
languages, because of their well-defined semantics,
and their focus on graphical and executional aspects,
are desirable but not yet sufficiently mature and
popular.
BP adopting in interoperational BPML category,
which isjust used for support of collaboration,
mostly emphasize on interactions instead of abstract
BP modeling from real world, also their flexibility
level and understandability problems for users are
serious concerns. Monitorial BP Languages are not
practically fitting in this context to evaluate, but
promising.
6 CONCLUSIONS
In our paper, we presented an analysis and
evaluation of the Business Process Modeling
Languages categories in support of Collaborative
Networks. We review their suitability for supporting
collaboration among enterprises.
To ensure a systematic and methodological
approach in our review process, we have reviewed
publications addressing categories of business
process modeling in the context of BPMLs. Then,
we have discussed different BP languages, and from
a language-independent perspective, we have
introduced our six categories of BPMLs.
Additionally, we have identified a set of criteria
required for adopting BPs among enterprises in CNs.
Based on these defined set of criteria, the six BPML
categories are further analysed, regarding how they
fulfil the collaboration requirements for CNs.
As we have employed a partially qualitative
analysis approach, our analysis is not fully objective.
Although, based on the results showed in table 2, the
most suitable categories of BPMLs, especially for
adoption in Virtual Organizations and VO Breeding
Environments, are represented.
The elaborated results achieved through our
evaluation of BPMLs in the context of CNs, indicate
that depending on the requirements, the domain
experts may preferably select BPMN or OWL-s for
the purpose of their BP integration.
ACKNOWLEDGEMENTS
This work is partially supported by the FP7 project
GLONET, funded by the European Commission.
REFERENCES
Afsarmanesh, H., Camarinha-Matos, L.M., Msanjila, S.S.,
2011. Models, Methodologies, and Tools Supporting
Establishment and Management of Second-Generation
VBEs. Systems, Man, and Cybernetics, IEEE
Transactions on, 41(5), pp. 692-710.
Camarinha-Matos, L.M., Afsarmanesh, H., 2008. On
Reference Models for Collaborative Networked
Organizations. International Journal of Production
Research, 46(9), pp.2453-2469.
Chung, L., do Prado Leite, J., 2009. On non-functional
requirements in software engineering. Conceptual
modeling: Foundations and applications, pp.363-379.
Curtis, B., Kellner, M.I., Over, J.1992. Process modeling.
Communications of the ACM, 35(9), pp.75-90.
Damodaran, S., 2004. B2B integration over the Internet
with XML: RosettaNet successes and challenges.
In Proceedings of the 13th international World Wide
Web conference, pp. 188-195.
Devlin, B. A., & Murphy, P. T., 1988. An architecture for
a business and information system. IBM systems
Journal, 27(1), pp.60-80.
De Nicola, A., Lezoche, M., Missikoff, M., 2007. An
Ontological Approach to Business Process Modeling.
In Proceedings of IICAI-07.
Georgakopoulos, D., Hornick, M., Sheth, A., 1995. An
Overview of Workflow Management: From Process
Modeling to Workflow Automation Infrastructure. In
Distributed and Parallel Databases, 3(2), pp. 119-
153.
Goldkuhl, G., & Lagsten, J., 2012. Different roles of
evaluation in information systems research. In IT
Artefact Design and Workpractice Intervention.
Green, P., Rosemann, M., Indulska, M., Manning, C.,
2007. ”Candidate Interoperability Standards: An
Ontological Overlap Analysis”, Data & Knowledge
Engineering, Vol.62, No.2, pp. 274-291.
Havey, M., 2009. Essential business process modeling.
O’Reilly Media,.
Indulska, M., zur Muehlen, M., and Recker, J.,
2009.“Measuring Method Complexity: The case of
Analysis and Evaluation of Business Process Modeling Adoption in Collaborative Networks
31

Business Process Modeling Notation”, Report 09-
03:viewed on December 2012 in
www.BPMcenter.org.
ISO, International Organization for Standardization,
2010E. ISO/IEC/IEEE24765, Systems and Software
Engineering Vocabulary.
Kiepuszewski, B., ter Hofstede, A.H.M, van der Aalst,
W.M.P., 2002. “Fundamentals of Control Flow in
Workflows” (Revised version). ,QUT Technical
report, FIT-TR-2002-03.
Ko, R.K., Lee, S.S., Lee, E.W., 2009. Business Process
Management (BPM) Standards: a Survey. Business
Process Management Journal, 15(5), pp.744-791.
Lee, J., Siau, K., Hong, S., 2003. Enterprise Integration
with ERP and EAI. In Communications of the
ACM, 46(2), pp.54-60
Lu, R., Sadiq, S., 2007. A Survey of Comparative
Business Process Modeling Approaches. In Business
Information Systems. Springer Berlin, pp. 82-94.
Mendling, J., Reijers, H., Cardoso, J., 2007. What makes
process models understandable?. 5
th
international
Conference BPM, Vol. 4714 Springer, pp.48-63.
Mendling, J., Reijers, H.A., van der Aalst, W.M., 2010.
Seven Process Modeling Guidelines
(7PMG). Information and Software Technology, 52(2),
pp.127-136.
Milanivic, N., Milic, B., Malek, M., 2008. Modeling
Business Process Availability. In Proc. IEEE congress
on services (SERVICES), pp.315-321.
Mili, H., Tremblay, G., Jaoude, G.B., Lefebvre, E.,
Elabed, L., Boussaidi, G.E., 2010. Business Process
Modeling Languages: Sorting Through The Alphabet
Soup. ACM Computing Surveys (CSUR), 43(1), 4.
OASIS, 2008. “Reference Architecture for service
Oriented Architecture version 1.0”, viewed on
December 2012 available in: http://docs.oasis-
open.org/soa-rm/soa-ra/v1.0/soa-ra-pr-01.pdf.
Papazoglou, M.P., van den Heuvel, W.J, 2006. Service-
Oriented Design and Development Methodology.
International Journal of Web Engineering and
Technology, Vol. 2, No.4.
Recker, J., & Mendling, J., 2006. “ On The Translation
between BPMN and BPEL: Conceptual Mismatch
between Process Modeling Languages”, In
Proceedings 18th International Conference of
Advanced Information Systems Engineering, pp. 521-
532.
Roser, S., Bauer, B., 2005. A Categorization of
Collaborative Business Process Modeling Techniques.
E-Commerce Technology Workshops, Seventh IEEE
International Conference pp. 43-51.
Russell, N., ter Hofstede, A.H.M., van der Aalst, W.M.P.,
Mulyar, N., 2006. Workflow Control-flow Patterns: A
Revised View”, BPM center report, BPM-06-29,
viewed in Dec, 2012 at www.BPM-Center.org
Russell, N.C., 2007. Foundations of Process-Aware
Information Systems, Ph.D. Dissertation, Queensland
University of Technology.
Schonenberg, H., Mans, R.S., Russell, N.C., Mulyar, N.A.,
van der Aalst, W.M.P., 2008. “Process flexibility: a
survey of contemporary approaches”, Advances in
Enterprise Engineering I, Lecture Notes in Business
Information Processing, vol. 10, Springer, pp. 16–30.
Sudhakar, G. P., 2012. A model of critical success factors
for software projects. In Journal of Enterprise
Information Management, 25(6), pp.537-558.
Ter Hofstede, A.H.M., van der Aalst, W.M.P., Adams, M.,
Russell, N., 2010.“Modern Business Process
Automation: YAWL and Its Support Environment”,
Springer.
Trkman, P., 2010. The Critical Success Factors of
Business Process Management. International Journal
of Information Management, 30(2), pp.125-134.
Van der Aalst, W.M, ter Hofstede, A., Weske, M., 2003.
Business process Management: A survey. Business
Process Management, Springer, LNCS 2678, pp.1-12.
Van der Aalst, W.M., 2009. ProcessAware Information
Systems: Design, Enactment, and Analysis. Wiley
Encyclopedia of Computer Science and Engineering.
Van der Aalst, W.M., Dustdar, S., 2012. Process Mining
Put into Context. In Internet Computing, IEEE,
16(1),
pp.82-86
W3C-OWL-S, 2004. “Semantic Markup for Web
Services”, in http://www.w3.org/Submission/OWL-S
viewed in December 2012
Weske, M., 2007. Business Process Management:
Concepts, Languages, and Architectures. Leipzig,
Alemania: Springer-Verlag Berlin, pp.5-25.
Workflow Management Coalition (wfmc), 1999.
erminology & Glossary (WFMC-TC-1011) issue 3.0,
Dec.2012, in standards of http://www.wfmc.org.
Wolf, C., Harmon, P., 2012, “ State of Business Process
Management”, December 2012: www.bptrends.com
Zdun, U., Hentrich, C., Van Der Aalst, W.M., 2006. A
Survey of Patterns for Service-Oriented Architectures.
In International Journal of Internet Protocol
Technology, 1(3), pp. 132-143.
Third International Symposium on Business Modeling and Software Design
32