Extending BPMN 2.0 With the Knowledge Dimension
Mariam Ben Hassen, Mohamed Turki and Faïez Gargouri
University of Sfax, ISIMS, MIRACL Laboratory - B.P. 242, 3021 Sfax, Tunisia
mariem.benhassen@isims.usf.tn, mohamed.turki@isetsf.rnu.tn, faiez.gargouri@isims.usf.tn
Keywords: Knowledge Management, Knowledge Identification, Sensitive Business Process Modeling, BPM4KI
Meta-Model, Ontologies, BPMN 2.0, Extension Mechanism.
Abstract: This paper introduces BPMN4KM, an extension of the most suitable business process modeling formalism
BPMN 2.0 for modeling knowledge dimension in Sensitive Business Processes (SBPs). The extension is
designed methodically by application of the extension mechanisms of BPMN 2.0. We aim at incorporating
relevant issues at the intersection of Knowledge Management (KM) and Business Process Modeling (BPM)
in order to enrich the graphical representation of SBPs and improve the localization and identification of
crucial knowledge mobilized and created by these processes.
1 INTRODUCTION
Sensitive Business Process (SBP) modeling has
become primary concern for any successful
organization to improve the management of their
individual and collective knowledge. A SBP is
characterized by a high number of critical activities
with intensive acquisition, sharing, storage and
(re)use of very specific knowledge «crucial
knowledge», high degree of internal/tacit knowledge
created and exchanged among experts (who carry
out actions with high levels of expertise, creativity
and innovation), diversity of information and
knowledge sources, high dynamic conversion of
knowledge and high degree of collaboration and
interactions (intra/inter-organizational) between a
wide range of agents/experts. Moreover, it is
typically an unstructured or semi-structured BP,
encompasses a highly dynamic complexity.
In order to enrich and improve the SBP
modeling, we have proposed, in previous work, a
conceptual specification of SBP organized in new
multi-perspective meta-model, entitled «BPM4KI:
Business Process Meta-Model for Knowledge
Identification» (Ben Hassen et al., 2017b; Ben
Hassen et al., 2017c). BPM4KI explicitly organizes
the key concepts and relationships that characterize a
SBP. It integrates all relevant perspectives/
dimensions relating to BPM-KM, i.e. the functional,
the organizational, the behavioral, the informational,
the intentional and the knowledge perspectives. In
this research work, we focus more on the
«Knowledge Dimension» which is not yet
explicited, fully supported and integrated within BPs
models and BPM approaches and formalisms.
However, while importance of knowledge
dimension is well recognized, there is no clear
theoretical background and successful practical
experiments of inclusion and implementation of this
dimension in BP/SBP models. In such languages as
IDEF0, IDEF3, GRAPES BM in GRADE tool , EPC
diagrams in ARIS tool (ARIS Expert Paper, 2007),
UML 2.0 Activity Diagram (OMG, 2011) and
BPMN 2.0 (OMG, 2013), data, information and
material flows are often represented in BP models
by the same symbols/artifacts and without any
unambiguous definitions of the concepts. At the
same time knowledge has poor or no modeling
capabilities in these formalisms. On the other hand,
knowledge modeling languages (KMDL (Gronau et
al., 2005; Arbeitsbericht, 2009), GPO-WM (Heisig,
2006), PROMOTE (Woitsch and Karagiannis, 2005)
and NKIP (Netto et al., 2013)) have shortcomings
concerning their ability to explicitly and fully
include the knowledge dimension within BPs
models as well as relevant issues at the intersection
of KM and BPM. They have limited process
perspective representation, i.e. they do not address
process logic to full extent and thus there is no
possibility to represent data and information. To
address this research gap, we propose to extend one
of the best known modeling formalism, the Business
Process Modeling Notation (BPMN) (OMG, 2013),
with the knowledge dimension in order to explicitly
111
incorporate all relevant aspects related to KM within
BPs models, and on the other hand, to enrich the
graphical representation of SBPs and improve the
localization and identification of crucial knowledge
mobilized and created by these processes. In fact,
BPMN 2.0 was selected as the most suitable BPM
notations for SBP representation, because addresses
the highest representation coverage of the set of
BPM4KI concepts and incorporates requirements for
SBP modeling better than other formalisms (Ben
Hassen et al., 2016a; Ben Hassen et al., 2017a).
Nevertheless, the main weaknesses identified in this
specification regards the knowledge dimension
modeling.
The present work presents BPMN4KM: a BPMN
2.0 extension, including all relevant aspects related
to knowledge dimension in SBP modeling. The
proposed extension is developed using the
extensibility mechanisms of BPMN (OMG, 2013).
The rest of the paper is structured as follows:
Section 2 presents BPMN 2.0 and related works
relevant to the research problem. Section 3 presents
the central concepts that describe the knowledge
dimension of SBP modeling. Section 4 presents the
proposed approach for extending BPMN 2.0 with
the knowledge dimension. Section 5 concludes the
paper and underlines some future research topics.
2 BACKGROUND AND RELATED
WORK
This section presents background research: section
2.1 describes BPMN as one of the most suitable
BPM notations; section 2.2 briefly present related
works relevant to the research problem.
2.1 BPMN 2.0
BPMN 2.0.2 stands for Business Process Model and
Notation (OMG, 2013). It is a graphical
representation for specifying BPs in a BP model, and
a standard for BP modeling notations. BPMN is
initiated as a standard BPM language for
conventional business, B2B and services process
modeling. It can be used within many methodologies
and for many purposes, from high-level descriptive
modeling to detailed modeling intended for process
execution providing a standardized bridge for the
gap between BP design and its implementation.
BPMN considers notational elements grouped in
five basic categories (Flow Objects, Data,
Connecting Objects, Swimlanes and Artifacts).
Besides, it has the capabilities of handling B2B BP
concepts, such as public, private, collaboration
processes and choreographies, as well as advanced
modeling concepts, such as exception handling and
transaction compensation in addition to the
traditional BP.
Several surveys have evaluated the adequacy of
BPMN for BPM. From our point of view, BPMN
has six main advantages (Ben Hassen et al., 2017):
It is the BPM standard backed up by OMG,
which is based upon a meta-model (OMG, 2013)
built with UML, the notation which is the de
facto standard for modeling software engineering
artifacts (OMG, 2007).
It is very simple, easy to use, readily
understandable and accessible by all business
stakeholders.
BPMN is one of the most recent and expressive
BPM notations, grounded on the experience of
earlier BPM formalisms, which ontologically
makes it one of the most complete BPM
formalisms (Recker et al., 2009).
It is appropriate for modeling collaborative BPs
actors that display complex flows with high
degree of interactions among process’ actors and
high degree of information exchanged,
developed and shared among participants.
It is currently the BP notation most used among
process modeling practitioners, with more BPM
tools support available.
BPMN is extensible. BPMN 2.0 defines an
extensibility mechanism for both process model
extensions and graphical extensions.
Finally, BPMN 2.0 presents the broadest
coverage of the set of BPM4KI meta-model
concepts (except the knowledge dimension) (Ben
Hassen et al., 2017a).
Based in the previous assessments, BPMN 2.0 is
taken as a basis for the representation of SBP
models.
2.2 BPMN 2.0 Shortcomings
BPMN stresses the process view representation,
offering a number of symbols for modeling various
decision points, process, activity and event types.
BPMN constructs emphasize mainly the support of
the control-flow and data perspective when
expressing processes’ orchestration and
collaboration. As other BPM formalisms, BPMN
constructs have a shallow coverage of informational,
organizational and intentional aspects of BPM.
Moreover, BPMN focuses entirely on the functional
and behavioral aspects of the BP model.
Seventh International Symposium on Business Modeling and Software Design
112
Nevertheless, the main weaknesses identified in this
specification regards the knowledge dimension
modeling which represents the core and relevant
dimension in SBP models (exploring the
collaboration and interaction aspects). Currently,
from the point of view of various ways how
knowledge (including data and information) are used
in organizations, the following issues are not yet
fully supported in BPMN 2.0 (neither in any of the
above-mentioned BPM and knowledge modeling
formalisms):
Opportunity to clearly distinguish between data,
information and knowledge in the representation
of flows between SBP activities. The information
and data exchange constitutes the basis for
knowledge dissemination and generation. Note
that, BPMN provides opportunity to model only
information and data flow using the same
symbols/artifacts and without any unambiguous
definitions of the concepts.
Opportunity to identify the different owners/
sources of knowledge involved in the BP
activities and location where knowledge can be
obtained and can be clearly stated.
Opportunity to consider the roles of humans in
BP activities, be it as humans (single persons),
teams, or communities of practice who bears the
internal/tacit knowledge.
Opportunity to integrate and separate the
different types /kinds of knowledge (tacit/explicit
dimension, internal/external dimension,
factual/procedural dimension,
individual/collective dimension, etc.).
Opportunity to integrate and separate the
different nature of knowledge (like experience,
basic knowledge, scientific/ technical
knowledge, general knowledge, etc.).
Possibility to illustrate knowledge flows between
sources and among activities.
Possibility to represent the dynamic of
acquisition, preservation, transfer, sharing,
development, and (re) use of individual and
organizational knowledge within and between
BPs activities.
Ability to specify more than two opportunities of
knowledge conversions (between knowledge
types) taking place in single SBP activity.
Opportunity to enable modeling the
critical/knowledge intensity dimensions of
organizational activities which are important to
determine the crucial knowledge mobilized and
created by these activities.
Opportunity to accurately represent collaborative
aspects and specify how do interactions occur
(information and knowledge exchange) in SBPs.
These aspects are useful to characterize the
SBPs, due to, for instance, the high degree of
knowledge exchanged and developed and shared
among agents through intra/inter-organizational
collaboration, and its dynamic nature. In fact,
BPMN 2.0 provides a specific choreography
model which allows to concentrate only on
conversation between performers. However, this
model does not show how performer’s
knowledge changes during the conversation and
communication.
To sum up, BPMN 2.0 diagrams are not
adequate for the new SBP modeling requirements.
So, to overcoming the discussed shortcomings,
BPMN 2.0 will be adapted and extended to be
convenient for a rich and expressive representation
of SBPs, including all or at least most of the relevant
issues at the intersection of KM and BPM.
2.3 Related Work
The integration of KM into BPs has rapidly become
the most promising practical and theoretical task in
KM. In this context, there have been several
attempts to integrate the knowledge
concept/dimension in BP models as well as in BPM
and knowledge modeling formalisms, e.g, (Gronau
et al., 2005; Woitsch et al.,2005; Weidong et al.,
2008; Supulniece et al., 2010; Businska et al., 2011;
Sulanow et al., 2012; Liu et al., 2012; Ammann et
al., 2010; Ammann et al., 2011; Ammann et al.,
2012; Netto et al., 2013; França et al., 2015; Gronau
et al., 2016).
However, none of the proposed knowledge
oriented BPM approaches and formalisms
adequately and fully support and represent all
relevant aspects of knowledge dimension within BPs
models (e.g., differentiation between tacit and
explicit knowledge, the different types of knowledge
conversion, the dynamic aspects of knowledge, the
different sources of knowledge, etc.). At the same
time, BPM is challenging - these notations are weak
in representing logic/ control flow of the BP and the
process perspectives as a whole (i.e., the structural,
behavioral, organizational and informational
dimensions).
Besides, while importance of knowledge
dimension is well recognized, there is no clear
theoretical background and successful practical
experiments of inclusion of this dimension in the
well-known BPM standard. In particular, there are
only a few initiatives in the BPM-KM area, which
use the BPMN as core formalism and systematically
Extending BPMN 2.0 With the Knowledge Dimension
113
enhance its capabilities and extend it by KM specific
aspects (Ammann et al. 2008, Ammann et al. 2012;
Ben Hassen et al., 2016). Ammann et al. (2008)
defined an extension of BPMN 1.1 (OMG, 2008) for
knowledge-related BPM, called BPMN-KEC (KEC
stands for knowledge, employees, and
communities). In this work different objects were
used: objects for knowledge and information, for
knowledge conversions, for associations and for
persons. Nevertheless, the proposal has not the
necessary expressivity and features to represent the
relevant SBP elements, including the knowledge
aspect. Another work by Supulniece et al. (2010),
proposed an extension of BPMN which roots in
concepts implemented in knowledge-oriented
modeling language (KMDL) (such as an information
object, knowledge object, type of knowledge
conversion) (Gronau et al., 2005) with few additions
and changes in graphical representation. However,
experiments with the integrated notation revealed
that the relationship between the phenomena behind
the symbols is somewhat unclear in the BPM.
Moreover, the relevant aspects of knowledge
dimension do not fully supported and represented
(like the different types of knowledge mobilized and
created by each BP activity, the knowledge flow, the
different sources/supports of knowledge, etc.).
To date, to the best of our knowledge, there is a
lack of works providing systematic approaches for
the development of extensions to the BPMN 2.0
meta-model to consider the knowledge aspect in
BPM. However, there are previous works providing
approaches to extend BPMN 2.0 (OMG, 2011) to
represent their domain specific requirements. Some
interesting extension proposals are presented in
(Charfi et al., 2011; Stroppi et al., 2011; Baumgrass
et al., 2014; Martinho et al., 2015; Jankovic et al.,
2015 Braun et al., 2015). The differences between
the different research works unveil the need for a
unified method for the conceptual modeling of
extensions and their representation in terms of the
BPMN extension mechanism.
In this paper, we aim to solve the discussed
shortcomings and address the gap between BPM and
KM. Precisely, this research work presents a
rigorous scientific approach to extend BPMN 2.0 for
KM. This extension must consider and incorporate
all relevant aspects of SBP modeling, including the
knowledge dimension, in order to allow a rich and
expressive representation of SBPs and improve the
localization and identification of crucial knowledge
mobilized and created by these processes.
3 MODELING SBPs: THE
BPM4KI META-MODEL
This section first introduces the notion of SBP and
then present an extract of BPM4KI, a BP
independent generic meta-model common to current
BPM formalisms which ensures the best suitability
to model SBP.
3.1 Notion of SBP
According to Ben Hassen et al., (2016b; 2017a), a
Sensitive Business Process is a BP which comprises
a high number of critical organizational activities
(individual/collective) with intensive acquisition,
sharing, storage and (re)use of very specific
knowledge « crucial knowledge». It mobilizes a
large diversity of information and knowledge
sources, consigning a great amount of heterogeneous
knowledge. Moreover, an SBP requires a high
dynamic conversion of knowledge and a high degree
of collaboration and interaction (intra/inter-
organizational) among participants. Its execution
involves many external agents and the assistance of
many experts, who apply, create and share a great
amount of very important tacit organizational
knowledge, in order to achieve collective objectives
and create value. In addition, SBP are typically an
unstructured or semi-structured organizational
actions, requires substantial flexibility,
encompassing a highly dynamic complexity. Due to
those characteristics, modeling and organizing the
knowledge involved in SBP is relatively critical.
3.2 BPM4KI: A BP Meta-Model for
Knowledge Identification
In order to enrich and improve the SBP modeling,
we proposed a semantically rich conceptualization
for specifying a SBP organized in a new generic
multi-perspective meta-model of BP representation,
the Business Process Meta-Model for Knowledge
Identification (BPM4KI). The enriched meta-model
serves two purposes: (i) to deepen the elements and
dimensions defining a SBP, by offering a coherent
conceptual specification for this BP type, and (ii) to
develop a rich and expressive graphical
representation of SBPs to improve the localization
and identification of crucial knowledge mobilized
and created by these processes. The current version
of BPM4KI offers a referential of generic concepts
and relationships relevant to the BPM-KM domain
semantically rich and well-based on «core» domain
Seventh International Symposium on Business Modeling and Software Design
114
Figure 1: An extract of BPM4KI meta-model: conceptual ontology design pattern relating to the knowledge
perspective/dimension of SBP modeling (with inter-aspects relationships).
Extending BPMN 2.0 With the Knowledge Dimension
115
ontologies (Kassel, 2005; Gangemi, 2006; Kassel,
2010; Kassel et l., 2012; Turki et al., 2016), which
are based on top of the DOLCE foundational
ontology (Masolo et al., 2004). BPM4KI were
categorized in six perspectives (or dimensions),
namely, the functional, the organizational, the
behavioral, the informational, the intentional and the
knowledge perspectives. The different dimensions
are crucial for a complete understanding,
characterization and representation of an SBP (Ben
Hassen et al., 2016b; 2017b; 2017c).
In this research work, we focus more on the
description and analysis of the knowledge dimension
which represents the most relevant aspects of SBP
modeling, exploring the KM aspect, the
collaboration and interaction and all relevant SBP
elements (such as individual and collective dimension
of activities; critical activities mobilizing crucial
knowledge; knowledge intensive activities; dynamic
aspects; collaboration and interaction among agents
contributing to knowledge creation and sharing;
The different aspects are required to characterize
the SBPs, due to the high degree of knowledge
exchanged and developed and shared among agents
through intra/inter-organizational collaboration and
to the frequent process evolution along time. We
point out that the knowledge dimension (supporting
the new SBP modeling requirements) is not yet,
however, not yet explicited, fully supported and
integrated within BPs models and BPM formalisms
(Ben Hassen et al., 2016b; 2017c). So, we aim at
obtaining new knowledge helpful for developing
BPM formalisms that could adequately support
above-mentioned issues in BP/SBP modeling.
The «Knowledge Perspective» is modeled as an
Ontological Design Patterns (ODP) (Gangemi et al.,
2006) represented as a UML class diagram. The
Knowledge ODP is based on the reuse and the
specification of central generic concepts (and the
relationships between them) defined in different
ontological modules of the global and consistent
ontology OntoSpec (Kassel, 2005; 2010): Action-
OS, Action of Organization-OS, COOP,
Partcipation-role-OS, Agentive Entity-OS,
Organization-OS, Function & Artefact-OS,
Capacity-OS, Artefact-OS, Resource-OS,
Communication-OS, I&DA-OS (Information and
Discourse Acts), IE&C-OS (Inscription, Expression
and Conceptualization) and Action Model-OS. These
ontological modules are available online
(http://home.mis.u-picardie.fr/~site-
ic/site/spip.php?article53), which are sufficient, on
the one hand, to broaden and deepen the knowledge
dimension elements, and on the other hand, to
characterize the useful concepts for a rigorous
specification and an enriched modeling of SBPs.
Figure 1 organizes and explicit the central concepts
of the knowledge perspective of BPM4KI (marked
in gray), in addition to inter-aspects relationships,
giving a view of all relevant aspects of the BPM4KI
meta-model as a whole.
According to Ben Hassen et al. (2016b), a
Knowledge is the Capacity (or disposition) to
perform (and affects) a type of Action aiming to
achieve an Objective. It isBorneBy an
Agentive Entity (as Human, Collective,
Expert or Organization). There are several
typologies of knowledge according to different
dimensions (Ben Hassen et al.. 2016b). For example,
Knowledge is divided into Internal
Knowledge, Explicit Knowledge and
External Knowledge according to the source
of knowledge dimension. Besides, Knowledge may
be either Propositional Knowledge or
Procedural Knowledge according to the
nature of knowledge dimension; Strategic
Knowledge and Familiarity Knowledge
according to the organizational value of knowledge
dimension. Moreover, knowledge can be divided
into Individual Knowledge and
Collective Knowledge according to the
organizational coverage of knowledge dimension.
With respect to the limited space of this paper, a
comprehensive description of the different concepts
present in this meta-model is detailed in (Ben
Hassen et al., 2016 b, Ben Hassen et al., 2017b; Ben
Hassen et al., 2017c).
Furthermore, it is important that an appropriate
BPM formalism provides explicit representation of
the different issues related to the knowledge
dimensions in BPM. In this context, the SBPs can be
graphically represented, using the well-known
standard for BPM, BPMN 2.0 (OMG, 2013), in
order to localize and identify the knowledge that is
mobilized and created by these processes. However,
as BPMN does not support the knowledge concept,
we have extended it. The following section explains
our extension proposal for including the knowledge
dimension in SBP modeling.
Seventh International Symposium on Business Modeling and Software Design
116
4 BPMN4KM: BPMN
EXTENSION FOR MODELING
THE KNOWLEDGE
DIMENSION
At the root of the success of modeling, design,
reengineering, and running BPs/SBPs is effective
use and support of organizational knowledge.
Knowledge must be considered as one of the BP
dimensions, because knowledge is related to action,
it is implemented in the action, and is essential to its
development. Knowledge is used to perform a
process, it is created as a result of process execution,
and it is distributed among process participants.
However, while importance of knowledge
dimension is well recognized, there is no clear
theoretical background and successful practical
experiments of inclusion, support and
implementation of this dimension in BP meta-
models and BPM approaches/formalisms (Ben
Hassen et al., 2017c). In this paper, we aim at
obtaining new knowledge helpful for developing
BPM formalisms that could handle all relevant
aspects related to knowledge dimension (including
data and information). Indeed, extending BP models
with the knowledge dimension would provide the
following benefits (Ben Hassen et al., 2017c):
Possibility to relate different forms of
knowledge, information and data to the BP
model.
Possibility to identify data, information and
knowledge inputs and outputs in different types
of organizational activities.
Illustrating and separating the data, information
and knowledge sources/owner that are required
to perform BP activities and knowledge that are
generated, created and/or modified as a results of
activities.
Enhance the localization of knowledge (where
knowledge can be obtained and clearly stated),
experts who hold the (internal) knowledge) as
well as their characterization.
Integration and distinction of different
knowledge types/nature.
Specifying the different opportunities of
knowledge conversion between knowledge types
(the dynamic sharing, dissemination, generation
and use of existing knowledge).
Possibility to represent knowledge flows
between sources, and among activities which are
about creation, organization, distribution and
reuse of knowledge among BP participants.
Giving an opportunity to improve understanding
about the knowledge usefulness, validity, and
relevance for particular activities (i.e. critical
activities) in an SBP.
Possibility to evaluate the amount of lost
knowledge if a person-owner of knowledge-
leaves the organization (to identify which tacit
knowledge in which cases should be transformed
into explicit knowledge).
According to the above-mentioned arguments
knowledge and BPs are directly related and their
integrated consideration is indispensable. In this
section, we propose a BPM technique that supports
an integrated consideration of BPs and knowledge.
The proposed technique is an extension of BPMN
2.0.2 (OMG, 2013), where the standard notation is
supplemented with knowledge modeling related
concepts. Despite its expressiveness, BPMN 2.0
does not yet explicitly represent the key concepts of
the Knowledge perspective (such as Individual
Tacit Knowledge, Collective Tacit
Knowledge, Expert, Explicit
Knowledge, External Knowledge,
Socialization, Externalization,
Internalization, etc.). To overcoming the
shortcomings of BPMN 2.0, some of its concepts
must be adapted and extended to include all or at
least most of the relevant SBP elements. In this
context, BPMN 2.0 defines four standard extension
mechanisms that are important for extending SBP
model with knowledge dimension. We have
introduced the main concepts of the knowledge
dimension into BPMN with a some additions and
changes in graphical representation.
4.1 The BPMN 2.0 Meta-Model
The BPMN formalism definition is based upon a
meta-model (OMG, 2013), which describee the
notation’s abstract syntax (by means of meta-classes,
meta-associations and cardinality constraints). The
BPMN meta-model includes elements from three
diagrams, targeting the following different purposes:
(i) for modeling processes’ orchestration and
collaboration diagrams; (ii) to simplify the
perspective of collaboration diagrams through
conversation diagrams and (iii) for modeling
participant’s interactions through the choreography
perspective. In this paper, from the full meta-model
that includes 151 meta-classes and 200 meta-
associations, we only consider the subset of elements
concerning the orchestration and collaboration
diagrams.
Extending BPMN 2.0 With the Knowledge Dimension
117
Main Concepts of BPMN 2.0. Meta-Model. The
OMG’s BPMN meta-model (OMG, 2013) considers
the four main dimensions of BPM:
The functional and behavioral dimensions of BPs
support the description of BP activities and their
synchronization along with events happening during
process execution through the notions of
FlowElementContainer (which can be either a
Process or a SubProcess) is a container of
instances of FlowElement. A flow element can be
either a FlowNode, a SequenceFlow or a
DataObject. A SequenceFlow is used to show
the order of various kinds of FlowNode elements
and the interactions between the participants. A
SequenceFlow may refer to an Expression
that acts as a gating condition. Instances of
SequenceFlow can link various kinds of
FlowNode elements. A FlowNode can be one of
the several different kinds of Activity, Event or
Gateway. A Gateway is used to control how
SequenceFlow interact within a process. An
Event is something that happens during the course
of a process. It can correspond to a trigger, which
means it reacts to something (catchEvent), or it
can throw a result (throwEvent). An Event can
be composed of one or more
EventDefinitions. There are many types of
Event Definitions:
ConditionalEventDefinition, Timer
EventDefinition, etc. An Activity is a
work performed within a process. An Activity
can be a Task (i.e. an atomic activity), a Sub
Process (i.e. a non-atomic activity) or a
CallActivity. A Task is used when the work
is elementary (i.e. it cannot be more refined).
BPMN2.0 identifies different types of tasks:
Service Task, User Task, Manual Task,
Send Task and Receive Task. The meta-class
Process describes a sequence of instances of
Activity carried out in an organization with
some specific objectives. If a process interacts with
other processes, it must participate in a
Collaboration. The collaboration is a way of
grouping several participants. Each Participant
(aka Pool) must address only one process. Given the
fact that a Participant is also an
InteractionNode, it can send or receive several
instances of MessageFlow.
Regarding the organizational dimension of
processes, an activity is accomplished by a
ResourceRole. A ResourceRole can refer to
a Resource. A Resource can define a set of
parameters called ResourceParameters. A
ResourceRole can be a Performer, which can
be a HumanPerformer, which can be in turn a
PotentialOwner. Besides, the LaneSets (i.e.
pools and lanes) allow grouping BPMN 2.0 model
elements according to participants of the process,
information systems, organization structure, etc.
Regarding the informational dimension of
processes, an ItemAwareElement references
element used to model the items (physical or
information items) that are created, manipulated and
used during a process execution. The
ItemAwareElement is an abstract meta-class,
from which derives several data related meta-classes
representing transient (DataObject) or persistent
(DataStore) data containers, as well as input or
output data to/from Activity by means of meta-
classes derived from Data Association. It
includes DataObject, Data Object
Collection, DataObjectReference,
Property, Data Store Artifact, Data
Input or Data Output (Collection). Moreover,
the Artefacts (i.e. Group and Annotation) allow
representing process data.
4.2 Mapping BPMN&BPM4KI
Meta-Models: Analysis of BPMN
Support for the Knowledge
Dimension Concepts
As shown in Table 1 BPMN lacks support for
several concepts of the knowledge aspect meta-
model (the ODP relating to the knowledge
perspective of SBP modeling). Therefore, to remedy
for this lacks, we define an extension of the BPMN
specification, called BPMN4KM, which introduces
the knowledge dimension aspects and provides a
rich and expressive representation of SBPs to
identify and localize the crucial knowledge
mobilized by these BPs.
In fact, we argue that an extension should widely
make use of standard elements in order to exhaust
the vocabulary of BPMN and reduce new elements
to a minimum. Based on both the specific SBP
domain concepts and requirements, the comparison
with standard BPMN is conducted in order to
identify a reasonable need for extension. According
to the presented knowledge ODP (Section 3.2), each
concept is examined regarding its semantically
equivalence with standard elements. Therefore, the
respective element descriptions, rules and
explanations within the BPMN specification (OMG,
2013) were analyzed in-depth. This leads implicitly
Seventh International Symposium on Business Modeling and Software Design
118
Table 1: Analysis of the BPMN support for the knowledge dimension ODP/ meta-model (with relevant inter-aspects
relationships) and derivation of concepts for the BPMN meta-model of the extension.
BPM4KI Concepts
Equivalence Check/BPMN Concept
Support Level
Knowledge Perspective
Knowledge
- (No equivalence)
-
Internal Knowledge
-
Tacit Knowledge
-
-
Latent Knowledge
-
-
Conscious Knowledge
-
-
Explicit Knowledge
-
-
External Knowledge
-
-
Procedural Knowledge
-
-
Propositional Knowledge
-
-
Strategic Knowledge
-
-
Familiarity Knowledge
-
-
Individual Knowledge
-
-
Collective Knowledge
-
-
Organizational Knowledge
-
-
Physical Knowledge Support
-
-
Behavioral
Perspective
Message Flow
Equivalence Message Flow
+
Association
Equivalence Association
+
Functional Perspective
Action Of Collective
Conditional equivalence ˡ Process
+
Organizational Activity
Equivalence Activity, Task, Sub Process
+
Deliberate Action
Conditional equivalence Activity
Partly
Discourse Act
Conditional equivalence Activity, Task
Partly
Critical Organizational Activity
Conditional equivalence Activity
Partly
Collaborative Organizational
Activity
Conditional equivalence Activity,
Choreography Activity
Partly
Knowledge Intensive Activity
Conditional equivalence Activity
Partly
Communicative Interaction
Conditional equivalence Activity,
Choreography, Collaboration, Conversation
-
Socialization
- (No equivalence)
-
Internalization
-
-
Explicitation
-
-
Externalization
-
-
Combination
-
-
Organizational
Perspective
Agentive Entity
Conditional equivalence Resource
Role/Performer, Participant (Partner/Role
Entity)
Partly
Collective
Conditional equivalence Resource
Role/Performer, Participant
Partly
Organization
Conditional equivalence Resource Role
Partly
Human
Equivalence Resource Role, Human
Performer
+
Experiencer
Conditional equivalence Human
Performer, Potential Owner
Partly
Expert
- (No equivalence)
-
Information
Perspective
Information
- (No equivalence)
-
Information Medium
Partly
Physical Artefact
Conditional equivalence Data Object
Partly
Data
- (No equivalence)
-
Discourse
-
-
Intentional
Perspective
Sensitive Business Process
Conditional equivalence Process
Partly
Distal Intention
- (No equivalence)
-
Objective
-
-
1
Process only define the Action of Organization (Business Process) which is an Action of Collective performed by a group of individuals
affiliated with the organization (Kassel et al., 2012). However, Process cannot be used to specify the actions that can be carried out
collectively by the individuals making up the Collective.
to the derivation of the BPMN4KM meta-model and
its stereotypes.
According to (Braun et al., 2015), the following
rules are defined for the equivalence check
Extending BPMN 2.0 With the Knowledge Dimension
119
(correspondence between concepts of the knowledge
perspective ODP/meta-model (extract of BPM4KI)
and the BPMN mata-model):
- Equivalence: There is a semantically equivalent
construct in the BPMN in the sense of a permitted
combination of elements or just a single element. In
this case, no extension is necessary and the domain
concept is represented as BPMN concept.
- Conditional equivalence: There is no obvious
semantic matching with standard elements, but
rather situational discussion is necessary in order to
provide arguments for a possible mapping or to
explain why it is not feasible. This situation is
caused by the partial under specification of BPMN
elements (OMG, 2013). Consequently, the concept
is either treated as equivalent concept or as non-
equivalent concept.
- No equivalence: There is no equivalence to any
standard element for three reasons: First, the entire
concept is missing. In this case, the domain concept
is represented as Extension Concept in the
BPMN4KM meta-model. Second, a relation between
two concepts is missing. Therefore, an association
between the affected concepts is constructed in the
BPMN4KM meta-model. Third, properties of a
concept are missing. Then, an owned property is
assigned to the element in the extended model. Table
1 provides the conducted equivalence check and its
implications for the extended BPMN meta-model.
As result of the correspondence check, the concepts
of the BPMN4KM meta-model are
classified/characterized as BPMN Concepts (are
those that match with some concept of the BPMN
meta-model) or as Extension Concepts (are those
defined in the domain of the extension).
The following section shows the developed
BPMN meta-model extension using the BPMN 2.0
extensibility mechanisms.
4.3 The BPMN4KM Meta-Model
The BPMN meta-model (OMG, 2013) can be
extended by integrating new domain-specific
concepts to standard and predefined BPMN
elements. This is supported by a standard extension
mechanism consisting of four elements:
ExtensionDefinition- specifies a named
group of new attributes, that can be used by
standard BPMN elements. Thus, both new
concepts and new additional attributes can be
defined (jointly added/attached to the original
BPMN elements).
ExtensionAttributeDefinition-
defines new /particular attributes that can be
specified for an ExtensionDefinition
element.
ExtensionAttributeValue - contains
the value assigned to an extension attribute of a
BPMN element.
Extension- binds/imports the entire
ExtensionDefinition element and its
attributes to a BPMN model definition in order to
make them technically accessible.
Figure 2 presents the Class Diagram of BPMN
extension. By associating a BPMN element with an
ExtensionDefinition, every BPMN element
which subclasses the BPMN BaseElement can
be extended with additional attributes. Therefore,
BPMN 2.0 with their different extension
mechanisms appear to provide the most complete
coverage of the concepts and constructs needed for
analyzing and modeling most of the SBP
characteristics.
Despite the fact that BPMN offers a well-defined
extension interface, only very few BPMN extensions
make use of it (Braun et al., 2014), what hampers
comprehensibility, comparability between developed
extensions and impedes the straightforward
integration of extensions in modeling tools. We
suppose, that the missing procedure model for
extension building in BPMN causes this lack of
rigor.
Extension
+mustUnderstand: Boolean
Definitions
BaseElement
+id: String
ExtensionDefinition
+name: String
ExtensionAttribute Value
ExtensionAttribute Definition
+name: String
+type: String
+isReference: Boolean
Element
+valueRef
0..1
*
+value
1
0..1
+extensions
*
+definition
1
1
+extensionDefinition
+extensionAttributeDefinitions
1
*
+extensionAttributeDefinition
1
+extensionValues
1
*
Figure 2: BPMN extension class diagram.
Based on the model transformation rules stated
in Stroppi et al. (2011), we define the BPMN4KM
extension model (BPMN+X model). Figure 3 below
presents the resulting extended BPMN meta-model.
In this figure only the relevant standard BPMN
classes are shown in white. The BPMN4KM
concepts are shown in grey. We associate
Seventh International Symposium on Business Modeling and Software Design
120
Figure 3: Abstract syntax of the BPMN4KM extension.
Extending BPMN 2.0 With the Knowledge Dimension
121
Knowledge concept with the RootElement of
the BPMN specification. The semantics and the
abstract syntax of the BPMN4KM elements are
based on the specification of the BPMN extension
mechanism (OMG, 2013). BPMNElement allows
representing an original element of the BPMN meta-
model. ExtensionElement allows representing a new
element in the extension model which is not defined
in the BPMN meta-model (such as Knowledge,
InternalKnowledge, TacitKnowledge,
ExplicitKnowledge,
ProceduralKnowledge,ExternalKnowlede
, PhysicalKnowledgeSupport,
Information, DistalIntention,
Combination, Socialization,
Internalization, Externalization and
Explicitation). ExtensionDefinition allows
specifying a named group of attributes which are
jointly added to the original BPMN elements (such
as KnowledgeFlow, Experiencer,
Collective, Information Medium,
KnowledgeIntensiveActivity,
CriticalOrganizationalActivity,
CollaborativeOrganizationalActivity,
and Sensitive Business Process).
ExtensionDefinition has the same meaning than the
ExtensionDefinition element of the BPMN
metamodel. The semantics defined by the
ExtensionAttributeDefinition element of the BPMN
meta-model is captured by the Property metaclass of
the UML metamodel. Thus,
ExtensionAttributeDefinition is represented in
BPMN4KM models by UML properties, either
owned by the ExtensionDefinition elements or
navigable from them through associations. The
properties of ExtensionDefinition and
ExtensionElement elements can be typed as a
BPMNElement, ExtensionElement, BPMNEnum,
ExtensionEnum or UML primitive type.
Finally, ExtensionRelationship specifies a
conceptual link between a BPMNElement and a
ExtensionDefinition element aimed to extend it. The
BPMN extension mechanism cannot express the
BPMN element to be extended by an extension
definition. Thus, the definition of an
ExtensionRelationship does not produce any effect
in the resulting BPMN extension.
ExtensionRelationship is provided to help
conceptualizing extensions since extensions are
generally defined to customize certain elements of
the BPMN meta-model.
With respect to the limited space of this paper,
the application of each applied transformation rule
cannot be presented.
5 CONCLUSION AND FUTURE
WORK
This research work presents BPMN4KM: a BPMN
extension to explicitly represent, integrate and
implement the knowledge dimension in BP/SBP
models. It allows a rich and expressive
representation of SBPs in order to improve the
localization and identification of crucial knowledge
mobilized and created by these processes. The
proposed approach extension is developed using the
extensibility mechanisms of BPMN.
Our current research activities focus on
achieving the implementation of the proposed
extension according to BPMN4KM meta-model.
As further work, we will validate the
BPMN4KM meta-model by instantiating it in depth
(using extended BPMN) with real medical care
processes in the context of the Association of
Protection of the Motor-disabled of Sfax-Tunisia
(ASHMS) (Ben Hassen et al., 2017a), in order to
verify the completeness of the proposed concepts.
Another issue we will address with BPM4KI and
BPMN4KM is to propose a solution to model and
specify SBPs integrating relevant aspects related to
all BPM4KI dimensions. The general framework we
will propose for supporting SBP representation
advocates a model driven engineering approach
considering at the CIM level, a specific meta-model,
the BPM4KI meta-model for modeling SBPs, and at
the PIM level, an extension of BPMN (BPMN4SBP
meta-model). We aim at automatically generating
SBP models to enhance the knowledge
identification.
REFERENCES
Ammann, E.,2008. BPMN-KEC – An Extension of
BPMN for Knowledge-Related Business Process
Modeling”,Internal Report, Reutlingen University.
Ammann,E., 2010. A hierarchical modelling approach to
intellectual capital development. In Electronic Journal
of Knowledge Management, vol. 8, Issue 2, C.
Bratianu, Ed, 2010, pp. 181–191.
Ammann, E., Navas-Delgado, I., & Aldana-Montes, J.
F.,2011. A Knowledge Development Conception and
its Implementation: Ontology Categories, Knowledge
Ontology, Rule System and Application Scenarios.
Seventh International Symposium on Business Modeling and Software Design
122
International Journal on Advances in Life Sciences,
vol 3 no 3 & 4.
Ammann, E., 2012. Modeling of Knowledge-Intensive
Business Processes. International Journal of Social,
Behavioral, Educational, Business and Industrial
Engineering Vol:6, n.11.
Arbeitsbericht.: KMDL® v2.2 (2009). Available at
http://www. kmdl.de/
ARIS Expert Paper (2007): Business Process Design as
the Basis for Compliance Management, Enterprise
Architecture and Business Rules.
Ben Hassen, M., Turki, M., Gargouri, F., (2016a).
Choosing a Sensitive Business Process Modeling
Formalism for Knowledge Identification. Procedia
Computer Science, vol 100 (2016a) 1002–1015
Ben Hassen, M., Turki, M., Gargouri, F., (2016b). A
Proposal to Model Knowledge Dimension in Sensitive
Business Processes. In Madureira A., Abraham A.,
Gamboa D., Novais P. (eds) Intelligent Systems
Design and Applications. Advances in Intelligent
Systems and Computing. Springer, vol 557, 1015-
1030
Ben Hassen, M., Turki, M., Gargouri, F., (2017a). Sensitive
Business Processes Representation: A Multi-
Dimensional Comparative Analysis of Business Process
Modeling Formalisms. In: International Symposium
on Business Modeling and Software Design. Revised
Selected Papers (Book Chapter). Lecture Notes in
Business Information Processing, 275, Springer 2017,
pp. 83-118
Ben Hassen, M., Turki, M., Gargouri, F., (2017b).
Functional Dimension representation in Sensitive
Business Process Models. In International Conference
on Knowledge Management, Information and
Knowledge Systems (KMIKS’2017), Hammamet-
Tunisie.
Ben Hassen, M., Turki, M., Gargouri, F., (2017c). Towards
Extending Business Process Modeling Formalisms with
Information and Knowledge Dimensions. In 30th
International Conference on Industrial and Engineering
Applications of Artificial Intelligence and Expert
Systems (IEA/AIE 2017), Arras, France.
Baumgrass, A., Herzberg, N., Meyer, A., & Weske, M.
(2014). BPMN extension for business process
monitoring. In EMISA, pp. 85-98.
Braun, R., Esswein, W., 2014. Classification of Domain-
Specific BPMN Extensions. In: Frank, U., Pastor, O.,
Loucopoulos, P., Petrounias, I. (Eds.) Lecture Notes in
Business Infor-mation Processing, 197, 42-57.
Braun, R., Schlieter, H., Burwitz, M., & Esswein, W.
(2015). Extending a Business Process Modeling
Language for Domain-Specific Adaptation in
Healthcare. In Wirtschaftsinformatik , pp. 468-481.
Charfi, A., Turki, S. H., Cha? bane, A., Witteborg, H., &
Bouaziz, R. (2011). A model-driven approach to
developing web service compositions based on
BPMN4SOA. International Journal of Reasoning-
Based Intelligent Systems, 3(3-4), 194-204.
Businska, L., & Kirikova, M., 2011. Knowledge
dimension in business process modeling. In Forum at
the Conference on Advanced Information Systems
Engineering (CAiSE) (pp. 186-201). Springer Berlin
Heidelberg.
Businska, L., Supulniece, I., Kirikova, M., 2011. On data,
information, and knowledge representation in business
process models. In: The 20th International Conference
on Information Systems Development (ISD 2011),
Edinburgh, Scotland. Springer .
dos Santos França, J. B., Netto, J. M., do ES Carvalho, J.,
Santoro, F. M., Baião, F. A., & Pimentel, M. (2015).
KIPO: the knowledge-intensive process ontology.
Software & Systems Modeling, 14(3), 1127-1157.
Gangemi A., 2006. Ontology Design Patterns: A primer,
with applications and perspectives. Tutorial on ODP,
Laboratory for Applied Ontology Institute of
Cognitive Sciences and Technology CNR, Rome, Italy
(2006)
Gronau, N., Korf, R. and Müller, C., 2005. KMDL-
Capturing, Analysing and Improving Knowledge-
Intensive Business Processes. Journal of Universal
Computer Science, vol. 11, no. 4, 452-472.
Gronau, N., Thim, C., Ullrich, A., Vladova, G., & Weber,
E. (2016). A Proposal to Model Knowledge in
Knowledge-Intensive Business Processes. BMSD.-
Rhodes, Greece, 16, 98-103.
Heisig, P., 2006. The GPO-WM® method for the
integration of knowledge management into business
processes. In International Conference on Knowledge
Management, Graz, Austria, 331–337.
Jankovic, M., Ljubicic, M., Anicic, N. and Marjanovic, Z.,
2015. Enhancing BPMN 2.0 informational perspective
to support interoperability for cross-organizational
business processes. Computer Science and
Information Systems, 12(3), pp.1101-1120.
Kassel, G., 2005. Integration of the DOLCE top-level
ontology into the OntoSpec methodology.
Kassel, G., 2010. A formal ontology of artefacts”. Applied
Ontology 5(3-4) pp.223-246
Kassel, G, Turki, M, Saad, I, Gargouri, F., 2012. From
collective actions to actions of organizations: an
ontological analysis”. In Symposium Understanding and
Modelling Collective Phenomena (UMoCop), University
of Birmingham, Birmingham, England.
Liu, D.R., Lai D. R., Liu C.H., Chih-Wei, L., 2012.
Modeling the knowledge-flow view for collaborative
knowledge support. J. Know. Based. Syst. 31, 41-54.
Martinho, R., Domingos, D., & Varajão, J. (2015).
CF4BPMN: A BPMN Extension for Controlled
Flexibility in Business Processes. Procedia Computer
Science, 64, 1232-1239.
Masolo, C., Vieu, L., Bottazzi, E., Catenacci, C., Ferrario, R.,
Gangemi, A., Guarino, N., 2004. Social roles and their
descriptions”, In Dubois D., Welty C. (eds.), Proc. of the
Ninth International Conference on the Principles of
Knowledge Representation and Reasoning, pp. 267-277
Netto, J.M, Franca, J. B. S., Baião, F.A., Santoro, F. M.,
2013. A Notation for Knowledge-Intensive Processes.
In IEEE 17th International Conference on Computer
Supported Cooperative Work in Design,Vol.1, 1–6.
Extending BPMN 2.0 With the Knowledge Dimension
123
OMG Final Adopted Specification.2008. Business Process
Modeling Notation Specification”,
http://www.omg.org/spec/BPMN/1.1/
OMG (2013). Business Process Model and Notation
(BPMN), Version 2.0.2
http://www.omg.org/spec/BPMN/2.0.2/pdf /
OMG (2011). Unified Modeling Language (UML).
Version 2.0. from http://www.uml.org/
Recker, J., Rosemann, M., Indulska, M., Green, P., 2009.
Business process modeling: a comparative analysis.
Journal of the Association for Information, 10, 333–
363.
Stroppi, L. J. R., Chiotti, O., Villarreal, P.D., 2011.
Extending BPMN 2.0: Method and Tool Support.
Business Process Model and Notation, 5973
Sultanow, E., Zhou, X., Gronau, N., 2012. Modeling of
Processes, Systems and Knowledge: a Multi-
Dimensional Comparison of 13 Chosen Methods.
International Review on Computers and Software, 7(6)
3309-3319.
Supulniece, I., Businska, L., Kirikova, M.(2010). Towards
extending BPMN with the knowledge dimension. In:
Bider, I., Halpin, T., Krogstie, J., Nurcan, S., Proper,
E., Schmidt, R., Ukor, R. (eds.) BPMDS and
EMMSAD 2010. Springer, Heidelberg. LNBIP. 50,
69–81.
Turki, M., Kassel, G., Saad, I., Gargouri, F., 2016. A Core
Ontology of Business Processes Based on
DOLCE”. Journal of Data Semantics, 5(3), pp.165-177
Weidong, Z., Weihui, D., 2008. Integrated Modeling of
Business Processes and Knowledge Flow Based on
RAD. In IEEE International Symposium on
Knowledge Acquisition and Modeling, Wuhan, China,
49–53
Woitsch, R. , Karagiannis, D., 2005. Process Oriented
Knowledge Management: A Service Based Approach.
Journal of universal computer science 11(4) 565-588.
Seventh International Symposium on Business Modeling and Software Design
124