Extending BPMN 2.0 Meta-models for Process Version Modelling
Imen Ben Said
1
, Mohamed Amine Chaâbane
1
, Eric Andonoff
2
and Rafik Bouaziz
1
1
MIRACL, University of Sfax,
Route de l’aéroport, BP 1088, 3018 Sfax, Tunisia
2
IRIT-UT1, 2 rue du Doyen Gabriel Marty, 31042 Toulouse Cedex, France
Keywords: Business Processes, Versions, BPMN 2.0.
Abstract: This paper introduces BPMN4V (BPMN for Versions), an extension of BPMN for modelling variability
(flexibility) of processes before their use in an organizational context or before their publication over the
cloud as services. More precisely, this paper motivates the importance of modelling variability of processes
using versions and introduces the versioning pattern to be used to reach this objective. It also presents
BPMN4V, giving provided extensions to BPMN2.0 meta model, both considering versions of intra and
inter-organizational processes. An example illustrating the instantiation of the proposed meta-model is given
for each kind of process.
1 INTRODUCTION
With the advent of cloud computing and the
emergence of BPaaS (Business Process as a Service)
(Wang et al., 2010), enterprises and organizations
outsource their processes onto the cloud. Thus, the
cloud is now viewed as an environment in which
several processes may be used by several enterprises
which share common infrastructures and services.
As a consequence the cloud is the support for
process discovery, which requires process analysis
abilities in order to decide to use a given process or
not (Aalst, 2013a). Of course, all the dimensions
(organizational, informational and behavioural) of a
process have to be considered for this decision-
making.
Among the different processes spread over the
cloud, some are close to each other: they can be
viewed as versions or variants of a same process.
Instead of discovering this variability by analyzing
the different dimensions of the considered processes
or by analysing the log files resulting from their
execution using mining techniques (Aalst, 2013b),
we propose in this paper to define it since the
process modelling step, i.e., before their outsourcing
onto the cloud. Indeed, discovering the variability is
a difficult task and the solutions of the literature are
very few and not really convincing -e.g. (Luengo
and Sepulveda, 2011).
Process variability, and more generally process-
adaptation issue, has been deeply investigated these
last years. Several typologies for classifying
variability capabilities of languages or notations
have also been introduced (Schonenberg et al.,
2008), (Nurcan, 2008), (Weber et al., 2008),
(Andonoff et al., 2013), and three main types of
variability are identified: variability by design (or
flexibility) for handling foreseen changes in
processes, variability by deviation for handling
occasional unforeseen changes and where the
differences with initial process are minimal, and
finally, variability by evolution for handling
unforeseen changes in processes, which require
occasional or permanent modifications in their
schemas. However, existing languages and notations
introduced for modelling this variability (Kradofler
and Grepper, 1999), (Lu and Shadiq, 2006), (Zhao
and Liu, 2007), (Hallerbach et al., 2008), (Lu et al.,
2009), (Hallerbach et al., 2010), (Chaâbane et al.,
2011), (Angles et al., 2013) are incomplete as either
they deal with only one type of variability and do
not address these three types of variability in a
coherent framework, or they mainly focus on the
behavioural dimension of processes, or they are too
specific with low degree chance to be used, or
finally they are dedicated to specific domains.
On the other hand, in a previous paper, we have
defended the importance of versions to deal with this
variability issue and more precisely to deal with the
different types of variability (Chaâbane et al., 2009).
As a consequence, and in order to define a notation
that will perhaps be used in the future, we propose to
384
Ben said I., Chaâbane M., Andonoff E. and Bouaziz R..
Extending BPMN 2.0 Meta-models for Process Version Modelling.
DOI: 10.5220/0004891103840393
In Proceedings of the 16th International Conference on Enterprise Information Systems (ICEIS-2014), pages 384-393
ISBN: 978-989-758-029-1
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
extend BPMN2.0 (OMG, 2011), which is considered
as the de-facto standard for modelling processes, in
order to make the modelling of this variability
possible. More precisely, we extend the BPMN2.0
meta model to support the modelling of process
versions, considering all the dimensions of processes
(i.e., organizational, informational and behavioural
dimensions) in the context of intra and inter-
organizational processes.
We defend the use of BPMN2.0, instead of EPC
(Scheer et al., 2005) or UML activity diagrams
(Engles et al., 2005) for instance, because it is a
promoted standard and it can be easily understood
by all business users: business analysts that design
processes, the technical developers that are
responsible of technical implementation of those
processes, and finally people (actors or process
owners) that manage and control those processes
(Weske, 2007).
As stated before, the paper contribution is an
extension of BPMN2.0 to support the modelling of
inter or intra-organizational process variability
trough versions. This extension is called BPMN4V
(BPMN for Versions).
The paper is organized as follows. Section 2
gives a brief state of the art about modelling
variability of processes. It then presents our
approach for modelling this variability and
introduces the versioning pattern we use for that.
Sections 3 and 4 are respectively dedicated to the
presentation of BPMN2.0 meta-model extensions for
modelling versions of intra-organizational processes
(private processes) and versions of inter-
organizational processes. In both sections, an
example illustrates the definition of process
versions. Finally, section 5 concludes the paper,
highlighting its contribution and giving directions
for future works.
2 MODELING PROCESS
VARIABILITY USING
VERSIONS
This section first introduces related works and then
defends and presents our versioning approach for
modelling process variability.
2.1 Related Works
Variability has been the focus of numerous works in
the BPM area during the last decade. As indicated
before, several typologies for classifying variability
capabilities of notations or systems have been
introduced these last years (Nurcan, 2008),
(Schonenberg et al., 2008), (Weber et al., 2008)
(Andonoff et al., 2013). These works distinguish
three types of process variability: variability by
design for handling foreseen changes in processes,
variability by deviation for handling occasional
unforeseen changes in processes, where the
differences with the initial process are minimal, and
finally, variability by evolution for handling
unforeseen changes in processes, which require
significant occasional or permanent changes. In
addition to the three types of variability, these works
also considers two main times for changing
processes:
design-time, which corresponds to expected
changes that can be foreseen; in this case
variability can be a priori defined as it is
possible to model the whole process at design-
time;
run-time, which corresponds to unexpected
changes that can not be foreseen; in this case,
variability cannot be a priori defined as the
process will be changed at run-time.
Several contributions have been proposed to
address the variability issue. For instance, the notion
of variant is introduced in (Hallerbach et al., 2008)
and (Hallerbach et al., 2010). A variant is an
adjustment of a basic process model to unexpected
requirements of process contexts appearing at run-
time. As a consequence, these work mainly deal
with variability by evolution, at run-time. On the
other hand, several version-based meta-models for
capturing process changes have been proposed in the
literature (Kradofler and Geppert, 1999), (Zhao and
Liu, 2007), (Chaâbane et al., 2010), (Chaâbane et al.,
2011). For instance (Zhao and Liu, 2007) proposed
to model versions of processes as direct graphs
named VPG (Version Preserving directed Graph).
The nodes of VPG are activities of the process while
the arcs of VPG define coordination between these
activities. (Zhao and Liu, 2007) also proposed
operations to modify the VPG: create node, delete
node, replace node, create arc, etc. However, this
work mainly focuses on the behavioural dimension
of processes, neglecting the informational and
organizational dimensions which are however
important to have a comprehensive view of
processes (Aalst et al., 2003). (Chaâbane et al.,
2010) compensated this weakness but the proposed
notation is too specific and has no chance to be used.
Another interesting work is the one of (Weber et
al., 2008). The authors introduced the notion of
change pattern to define process changes. They
ExtendingBPMN2.0Meta-modelsforProcessVersionModelling
385
distinguish patterns to deal with variability at
design-time and run-time. However, again the
notation is too specific, too technical, and thus has
low chance of being used.
2.2 Our Approach for Process
Variability Modelling
We consider that the notion of version subsumes the
notion of variant. As illustrated in Figure1 below, a
version corresponds to one of the significant states a
process may have during its life cycle. So, it is
possible to describe the variability of a process
through its different versions. These versions are
linked by a derivation link; they form a version
derivation hierarchy. When created, a process is
described by only one version. The definition of
every new version is done by derivation from a
previous one: such versions are called derived
versions. Of course, several versions may be derived
from the same previous one: they are called
alternatives or variants. The derivation hierarchy
looks like a tree if only one version is directly
created from a process entity, and it looks like a
forest if several versions are directly created from
the considered process entity.
Figure 1: Versions to Model Process Variability.
To sum up, when considering versions, we
model both process evolution and process alternative
(i.e., variant) to describe process variability.
2.3 Versioning Pattern
This paper defends the idea of using versions to
make BPMN2.0 notation multi-versionable. Thus we
have defined a versioning pattern to support versions
for the main concepts of BPMN.
The versioning pattern we propose is very
simple: it includes only two classes: “Versionable”
class and “Version of Versionable” class, and two
relationships: “Is_version_of” and “Derived_From
as illustrated in Figure 2. A versionable class is a
class for which we would like to handle versions. In
addition we define a new class which contains
versions, called “Version of Versionable”.
Figure 2: Versioning Pattern.
The “Is_version_of” relationship links a class to
its corresponding versions. The “Derived_From
relationship allows for building version derivation
hierarchies (cf. Figure 1). This latter relationship is
reflexive and the semantic of both relationship sides
is the following: (i) a version (SV) succeeds another
one in the derivation hierarchy, and (ii) a version
(PV) precedes another one in the derivation
hierarchy. Regarding properties of a “Version of
Versionable” class, we introduce the classical
version properties, i.e., version number, creator
name, creation date and status.
3 MODELING VERSIONS OF
INTRA-ORGANIZATIONAL
PROCESSES
BPMN2.0 allows the creation of three basic types of
sub-models within an end-to-end process (OMG,
2011):
Process (private and public): a private process
is internal to a specific organization with the
objective of carrying out work. As for a public
process, it represents the interactions between
a private process and another process or
business entity.
Collaboration: a collaboration depicts the
interactions between two or more private
processes.
Choreography: a choreography is an extended
type of collaboration that defines the sequence
of interaction between business entities.
Intra-organizational processes can be modelled
using BPMN2.0 through private process concepts. A
private process describes a sequence of activities
performed within an organization in order to carry
out an objective. It is depicted as a directed graph.
This section first introduces BPMN concepts for
modelling private processes and then illustrates how
we use the previous versioning pattern to make some
of these concepts versionable.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
386
3.1 Concepts to Model Private
Processes
Figure 3 below introduces BPMN 2.0 meta-model
concepts useful to model private processes
(OMG,
2011).
Regarding the behavioural dimension of
processes, FlowElementContainer is an abstract
class used to define the superset of elements in
BPMN diagrams. An object FlowElementContainer
contains SequenceFlows, FlowNodes (Gateways,
Events, and Activities), and Data Objects. A
SequenceFlow is used to show the order of
FlowNodes in a process. A Gateway is used to
control how SequenceFlows interact within a
process. An Event is something that “happens”
during the course of a process. It affects the flow of
the process and usually has a cause or an impact, and
in general requires or allows a reaction. More
precisely, an Event can catch a trigger “CatchEvent
which means it reacts to something, or it can throw a
result “ThrowEvent”. An Event can be composed of
one or more EventDefinitions. An EventDefinition
refers to the triggers of CatchEvents and to the
Results of ThrowEvents. There are many types of
Event Definitions: ConditionalEventDefinition,
TimerEventDefinition, etc. An Activity is a work
performed within a process. An Activity can be a
Task (i.e., an atomic activity) or a Sub Process (i.e.,
a non-atomic activity). A Sub Process is refined via
Activities, Gateways, Events, and SequenceFlows. A
Task is used when the work is elementary (i.e., it
cannot be more refined). BPMN2.0 identifies
different types of Tasks such as Service Task, User
Task, Manual Task, Send Task and Receive Task.
Regarding the organizational dimension of
processes, an activity is accomplished by
ResourceRoles. A ResourceRole can refer to a
Resource. A Resource can define a set of parameter
called ResourceParameters. A ResourceRole can be
a Performer, which can be a HumanPerformer,
which can be in turn a PotentialOwner. The
Performer class defines the resource that will
perform or will be responsible for an Activity. A
Performer can be human (HumanPerformer), i.e., it
defines people that are assigned to Activities. A
PotentialOwner is a specialization of
HumanPerformer, used to define persons who can
claim and work on User Task.
Regarding the informational dimension of
processes, ItemAwareElement references elements
used to model the items (physical or information
items) that are created, manipulated and used during
a process execution. An ItemAwareElement can be a
DataObject, a DataObjectReference, a Property, a
DataStore, a DataInput or a DataOutput.
DataObject, Property and DataObjectReference
are ways for modelling data of each process.
DataObject elements have a graphic representation
on a process diagram, while
Properties have not.
DataObjectReference allows the reuse of a
DataObject in the same diagram. DataStore
provides a mechanism for Activities to retrieve or
update stored information used in a process.
DataInput and DataOutput are used to model data
needed and produced during an execution or as a
result of execution. Data requirements are captured
as DataInputs and InputSets. Data that are produced
are captured using DataOutputs and OutputSets.
These elements are aggregated in the
InputOutputSpecification class.
Figure 3: Extract of BPMN2.0 Meta-Model for Modelling Private Processes.
ExtendingBPMN2.0Meta-modelsforProcessVersionModelling
387
3.2 Extending BPMN Meta Model for
Modelling Versions of Private
Processes
This section explains the proposed BPMN
extensions to consider versions of processes. The
idea is to use the versioning pattern introduced
before (cf. Figure 2) to make some classes of the
BPMN meta-model versionable. Figure 4 below
presents the resulting meta model.
We propose to handle versions for six classes:
Process, Sub Process, Event, Activity,
ItemAwareElement and Resource, in order to take
into account process variability. This variability is
expressed by the different versions of these class
instances, each one representing a significant state
(alternative, i.e., variant, or derived) of the
considered element (e.g., a process). A new version
of an element (e.g., a process or a resource) is
defined according to the changes occurring to it:
these changes may correspond to the adding of a
new information (property or relationship) or to the
modification or the deletion of an existing one.
Actually, these changes can affect all the dimensions
of a process: (i) its behavioural dimension to
redefine how the process is achieved (process, sub
process, activity and event), (ii) its organizational
dimension to redefine the roles invoked by the
process (Resource) or, (iii) its informational
dimension to redefine the used or produced
information (ItemAwareElement). The general idea
is to keep track of changes occurring to elements
participating to the description of the way business
is carried out.
Regarding the organizational dimension of a
process, we create a new version of Resource when
we change its parameters. For instance, a Manager
resource may be defined using two parameters:
name and experience. A new version of Manager
may be defined if it becomes necessary to consider
another parameter (e.g. region of the manager) and
this definition can lead to the definition of a new
process in which this resource is involved. We also
create versions of Resource when there is a change
in its privileges. For instance, an Employee is a
HumanPerformer resource that performs three
activities. Because some activities of the process in
which this employee is involved become automatic,
the employee can perform anymore only two
activities. A new version of the employee has then to
be defined.
Regarding the informational dimension of
processes, and more particularly ItemAwareElement,
we consider that changes in the structure and/or the
type of an ItemDefinition results in the creation of a
new version. For example, if Report is an
ItemAwareElement corresponding to a paper
document (Itemkind is a Physical data), and if after
technical changes it becomes an electronic document
(Itemkind becomes an
Information data), then a new
version of Report has to be created.
Regarding the behavioural dimension of
processes, several classes can gather versions: Event,
Activity, Sub Process and, of course, Process. More
precisely regarding activities, we create a new
Figure 4: Extending BPMN2.0 Meta-model for Modelling Versions of Private Processes.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
388
version of an activity when there are changes in the
type of the activity (a manual activity becomes a
service one), in the involved resources, or in the
required or produced data. Regarding events, we
create a new version of an event when there is
change in the associated EventDefinition. For
instance, if an Alert is a signal event (i.e., it has a
SignalEventDefinition), and if after technical
changes it becomes a message event (i.e., it has a
MessageEventDefinition), then a new version of
Alert has to be created. Regarding sub processes and
processes, we create new versions when there are
changes in the involved activities or in the way they
are linked together (used patterns are changed).
On the other hand, BPMN meta-model provides
extension mechanisms through classes Extension,
ExtensionDefinition and
ExtensionAttributesDefinition and each proposed
extension should be assigned to these classes (OMG,
2011). In our case, we propose to add the
VersionExtensionDefinition class which is an
abstract super class for all versions of versionable
classes. The VersionExtensionDefinition class has a
set of attributes described into
VersionExtensionAttribute class. Actually, a version
class contains specific attributes such as version
number, creator name, creation date and status. Each
of Version of Versionable class introduced to model
process variability is a sub-class of the
VersionExtensionDefinition class.
3.3 Example
We illustrate in Figure 6 the instantiation of our
meta-model according to the damage compensation
process of an insurance company. This process is
shown in Figure 5. Basically, it contains five
activities: Receive Request, Review Request, Send
reject Letter, Calculate claim amount and Financial
settlement. For simplification reasons, only the
behavioural dimension of the process is presented.
The first version of this process is represented in
the left part of Figure 5. This version starts when
receiving a claim. The checking of the claim can be
hold only if Conditions of Review is satisfied.
Conditions of Review is a conditional intermediate
event that refers to the condition: the claim date
must not exceed 7 days from the accident date. After
checking the claim, a reject letter is sent if the
request is not accepted. Otherwise, the claim amount
is calculated (by the insurance manager), and the
financial service prepares the financial settlement.
The insurance company has modelled a second
version of this process to take into account a new
law that imposes the insurance companies to make
an expertise when the damage amount exceeds
1000$. The right part of Figure 5 illustrates this
second version of the process introducing the
Expertise activity and both modifying the
intermediate Conditions of Review event and the
Financial settlement activity (their type have
changed). To sum up, the variability of the damage
compensation process is defined by two versions of
the process itself, two versions of Conditions of
Review event and two versions of the Financial
settlement activity: the first version of this activity
holds for the first version of the process while the
second one holds for the second version of the
process. In addition, the sequence flows and patterns
have been modified in the second version of the
process.
Figure 5: Variability of the Damage Compensation
Process.
Regarding variability of the Conditions of
Review event, in the first version of the process, it is
a conditional event referring to the condition: the
claim date must not exceed 7 days from the accident
date (i.e., it has a ConditionnalEventDefinition),
while in the second process version, we add another
EventDefinition the TimerEventDefinition which
indicates that the checking of the claim have to be
hold in the beginning of the week (i.e., in Monday,
Tuesday or Wednesday). As a consequence, the
Conditions of Review event becomes a multiple
parallel event that has two EventDefinitions:
Conditional EventDefinition (the claim date must
not exceed 7 days from the accident date) and
TimerEventDefinition (claim checking have to be
hold in the beginning of the week).
ExtendingBPMN2.0Meta-modelsforProcessVersionModelling
389
Figure 6: Extract of BPMN4V Meta-Model Instantiation
according to the Damage Compensation Process.
4 MODELING VERSIONS OF
INTER-ORGANIZATIONAL
PROCESSES
An inter-organisational process is a collaboration
between organisations (each one represented by a
process) in order to carry out a common business
target. BPMN2.0 defines collaboration diagrams for
modelling processes crossing organisations
boundaries. These diagrams include an explicit
representation of permanent interactions between the
involved entities. These interactions are defined as
message flows, i.e., messages exchanged between
involved entities (OMG, 2011). Note that BPMN
collaboration diagrams are used to model tight inter-
organisational processes, in which the collaboration
infrastructure is well-established and the involved
partners are known before process execution
(Divitini et al., 2001).
We propose to introduce the notion of version in
such diagrams to deal with variability of inter-
organizational processes. As a consequence, this
section first introduces BPMN concepts for
modelling collaboration diagrams and then details
how we extend such diagrams to represent versions
of inter-organizational processes.
4.1 BPMN Meta-model for Modelling
Collaboration Diagrams
Figure 7 below gives the main BPMN concepts for
modelling collaboration.
Figure 7: Extract of the BPMN2.0 Collaboration Meta-
Model (OMG, 2011).
A Participant represents partner entities (e.g., a
company) and/or partner roles (e.g., a buyer, seller,
or manufacturer) that are involved in a collaboration.
A Participant is often responsible for the execution
of a Private process. During a collaboration,
participants are prepared to send and receive
Messages. A MessageFlow illustrates the flow of
messages between two InteractionNodes. An
InteractionNode element is used to provide a single
element as the source or the target of a
MessageFlow. An InteractionNode can be a
Participant, a Task or an Event.
4.2 Extending BPMN Meta-model for
Modelling Versions of
Collaboration Diagrams
We propose to use the versioning pattern introduced
before to make some classes of the previous meta-
model versionable. Figure 8 below presents the
resulting meta-model.
We propose to handle versions for only two
classes: Collaboration and Message, in order to take
into account inter-organizational process variability.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
390
A version of collaboration contains a set of
participants, each one refering to a version of a
process. It also contains a set of messages flow, each
one refering to a version of message. For the same
reason than the one presented in section 3.2, Version
of Message and Version of Collaboration are sub-
classes of the class VersionExtensionDefinition.
Figure 8: BPMN2.0 Meta-Model for Modelling Versions
of Collaboration.
As explained before, a new version of an element
(e.g., a collaboration or a message) is defined
according to changes occurring to it: these changes
may correspond to the addition of information
(property or relationship) or to the modification or
the deletion of an existing one. A new version of a
collaboration may also result from changes of
participants. When we add or delete a participant (of
a process involved in the collaboration), it is
necessary to adapt the current process to this change:
we have to integrate the added participant or to
overcome the absence of the deleted one.
A new version of a collaboration may also result
from message changes. Exchanged messages have
an important impact in collaborations flow. Thus,
any change in a sent or a received message affects
the involved activities, and as a consequence, the
involved process. So, when we add (or delete) a
message, we have to add (or to delete) a received
and a send activity, which leads to change the
process schema. Moreover, when we change the
structure or the kind of ItemDefinition referenced by
an existing message, we also have to change the
involved sent and received activities.
Finally, a new version of a collaboration may
result from a process change. Processes that
participate in a collaboration can change their
schema. A process can for instance add or delete
activities or change the pattern linking these
activities. In this case, the other processes involved
in the collaboration have in turn to adapt to this
change in order to go on ensuring the collaboration.
Figure 9 below summarizes all the situations that
lead to change a collaboration diagram.
Figure 9: Causes of Collaboration Changes.
4.3 Example
We illustrate in Figure 11 the instantiation of our
meta-model using the damage compensation inter-
organizational process of the insurance company.
Figure 10 gives the third version of the damage
compensation process, which extends the example
introduced in section 3.3. In this version, the
insurance company decides to modify its work
strategy subcontracting the expertise activity. Thus,
a collaboration between the insurance company and
an expert agency is required. The collaboration
initiates when a claimant sends a request to the
insurance company. After checking, the company
sends a reject to the claimant if the request is not
accepted. Otherwise, in case where the damage
amount exceeds 1000$, an expertise request is sent
to an expert. The expert proceeds to the expertise
and sends back a report that resumes the expertise
results. Based on the received report, the claim
amount is calculated and a financial settlement is
done.
ExtendingBPMN2.0Meta-modelsforProcessVersionModelling
391
Figure 10: Collaboration between the Damage
Compensation Process and the Expert Process.
Figure 11: Extract of BPMN4V Meta-Model instantiation
according to the Collaboration Example.
5 CONCLUSIONS
This paper has proposed a solution to model
variability of intra and inter organizational processes
using versions. More precisely, it has proposed
BPMN4V, an extension of BPMN2.0 to consider
versions of private processes (intra-organizational
processes) and versions of collaborations (inter-
organizational processes). For each of these kinds of
processes, this paper has extended BPMN2.0 meta-
models and argued why and when creating versions
of elements.
Our future works will take two directions. On the
one hand, we will implement the BPMN4V in order
to obtain a specific tool for modelling versions of
private and collaboration processes. The
consequence will be the introduction of a specific
graphical notation for versions. On the other hand,
we will investigate context of intra and inter
organizational processes. Our objective is to give
information about (i) version of process goals
(Korherr and List, 2006) and (ii) version of process
use, i.e., in which circumstances (which situations)
do we use a specific version of process instead of
another one (Saidani and Nurcan, 2009), (Chaâbane
et al., 2010). To sum up, we intend to extend
BPMN2.0 in order to incorporate the context
dimension.
REFERENCES
Aalst, W. M. P. et al., 2003. Advanced topics in workflow
management: issues, requirements and solutions.
Journal of Integrated Design and Process Science,
7(3), pp. 47-77.
Aalst, W. M. P., 2004. Business Process Management
Demystified: A Tutorial on Models, Systems and
Standards for Workflow Management. Lectures on
Concurrency and Petri Nets, Advances in Petri-Net,
Eichstätt, Germany, pp. 1–65.
Aalst, W. M. P., 2013a. Business Process Management: A
Comprehensive Survey. Software Engineering (to
appear).
Aalst, W. M. P., 2013b. Challenges in Service Mining:
Record, Check, Discover. International Conference on
Web Engineering, Aalborg, Denmark, July 2013, pp.
1–4.
Andonoff, E., et al, 2013. Adaptation des processus
d’entreprise. Chapter 3, in L’adaptation dans tous ses
états, Lopisteguy, P., Rieu, D. Roose, P. Edition,
Cepadues Editor.
Angles, R., et al., 2013. V-BPMI: a Variability-oriented
Framework for Web-based Business Process
Modelling and Implementation. IEEE International
Conference on Research and Challenges on
Information Science, Paris, June 2013, pp. 313–324.
Chaâbane, M. A., et al., 2009. Versions to Address
Business Process Flexibility Issue. International
Conference on Advances on Database and
Information systems, Riga, Latvia, september 2009,
pp. 2–14.
Chaâbane, M. A., et al., 2010. Modélisation
multidimensionnelle des versions de processus.
Journal on Ingénierie des Systèmes d’Information,
15(5), December 2010, pp. 89–114.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
392
Chaâbane, M. A., et al., 2011. Towards a Business Process
Versions Query and Manipulation Language.
International Conference on Information Technology
and E-Services, Sousse, Tunisia, August 2011, pp. 12–
19.
Divitini M., Hanachi C., Sibertin-Blanc C., «Inter-
Organizational Workflow for Enterprise Coordination.
Coordination of Internet Agents», chapter 15,
Tolksdorf Editors. Springer Verlag, 2001.
Engles, G., et al., 2005. Process Modelling using UML.
Process-aware Information Systems: Bridging People
and Software through Process Technology, Chapter 5,
Wiley and Sons Editor, pp. 85–117.
Hallerbach, A. et al., 2008. Managing Process Variants in
the Process Life Cycle. International Conference on
Enterprise Information Systems, Barcelona, Spain, pp.
154–161.
Hallerbach, A. et al., 2010. Capturing Variability in
Business Process Models: the Provop Approach.
Journal of Software Maintenance 22(6-7), pp. 519–
546.
Korherr, B., List, B., 2006. Extending the UML 2 Activity
Diagram with Business Process Goals and
Performance Measures and the Mapping to BPEL,
Proceedings of the 2nd International Workshop on
Best Practices of UML, International Conference on
Entity Relationship, Tucson, Arizona, USA, Springer
Verlag, November 2006, pp. 7–18.
Kradofler, M. and Geppert, A., 1999. Dynamic Workflow
Schema Evolution based on Workflow Type
Versioning and Workflow Migration. International
Conference on Cooperative Information Systems.
Edinburgh, Scotland, pp. 104–114.
Lu, R. et al., 2009. Defining Adaptation Constraints for
Business Process Variant. International Conference on
Business Information Systems, Poznan, Poland, April
2009, pp. 145–156.
Lu, R., and Sadiq, S., 2006. Managing Process Variants as
an Information Resource. International Conference on
Business Process Management, Vienna, Austria,
September 2006, pp. 426–431.
Luengo, D. and Sepulveda, M., 2011. Applying Clustering
in Process Mining to find different versions of a
business process that changes over time. Business
Process Intelligence Workshop at International
Conference on Business Process Management,
Clermont-Ferrand, France, September 2011, pp. 153–
158.
Nurcan, S., 2008. A Survey on the Flexibility
Requirements related to Business Process and
Modelling Artefacts. International Conference on
System Sciences, Waikoloa, Big Island, Hawaii, USA,
pp. 378–387.
OMG, 2011. Business Process Model and Notation
(BPMN) Version 2.0. OMG Document Number:
formal/2011-01-03, Standard document URL:
http://www.omg.org/spec/BPMN/2.0, January 2011.
Saidani, O. and Nurcan, S., 2009. Context-Awareness for
Adequate Business Process Modelling. IEEE
International Conference on Research Challenges in
Information Science, Fés, Morocco, pp. 177–186.
Scheer, A.W., et al., 2005. Process Modelling using
Event-Driven Process Chain. Process-aware
Information Systems: Bridging People and Software
through Process Technology, Chapter 6, Wiley and
Sons Editor, pp. 119–145.
Schonenberg, H., et al., 2008. Process Flexibility: A
Survey of Contemporary Approaches. In the
International Workshop on CIAO/EOMAS at
International Conference on Advanced Information
Systems. Montpellier, France, pp. 16–30.
Wang M., et al., 2010. Process as a service. International
Conference on Services Computing, Miami, Florida,
July 2010, pp. 578–585.
Weber, B., et al., 2008. Change Patterns and Change
Support Features - Enhancing Flexibility in Process-
Aware Information Systems. Journal on Data and
Knowledge Engineering, 66 (3), pp. 438–466.
Weske M., 2007. Business Process Management:
Concepts, Languages, Architectures. Springer-Verlag
Editor, 2007.
Zhao, X. and Liu, C., 2007. Version Management in the
Business Change Context. International Conference
on Business Process Management, Brisbane,
Australia, September 2007, pp. 198–213.
ExtendingBPMN2.0Meta-modelsforProcessVersionModelling
393