Transformation of BPMN Model into an OWL2 Ontology
Mariem Kchaou
1
, Wiem Khlif
1
, Faiez Gargouri
1
and Mariem Mahfoudh
1,2
1
Mir@cl Laboratory, University of Sfax, Sfax, Tunisia
2
University of Kairouan, Kairouan, Tunisia
Keywords: BPMN Model, Ontology, OWL2, Transformation Rules, Business Context.
Abstract: Each enterprise needs to have a clear vision of its business processes in order to increase the quality of its
products/services. To fulfil this need, many enterprises rely on an Information System (IS). Most of the
previous systems were previously framed by applying business process model. In addition, the current trend
expresses a growing demand of reusing data from older information systems, which is very beneficial for the
implementation of semantic knowledge. The transformation of a BPMN model into an ontology leads to
reduce cost by reusing older systems. Although many studies are elaborated for transforming BPMN model
into ontology, they have not fully proposed the transformation rules. This paper suggests the addition of rules
for transforming annotated BPMN models to ontologies by accounting for the semantics of the BPMN model,
and providing for all business objects and activities. In addition, the transformations have the merit of
generating the OWL2 graphical representation.
1 INTRODUCTION
An automated information system (IS) gives
important support to the business process if its
capacities are best exploited. Indeed, the crucial
advantage of designing system by utilizing BPMN
notation lies in its ability to describe, and reflect the
real world of information systems better.
Furthermore, it has also got the further support of
developers. Hence, the BPMN notation has gradually
gained its popularity. However, the specification is
comprehensive and partially conflicting. Therefore,
several researches present an ontology that provides
a formal definition of BPMN and can be used as a
knowledge base (Annane et al., 2019). It is a formal
representation of knowledge and consists of
statements that define concepts, relationships, and
constraints. According to (Noy and McGuinness,
2001), an ontology allows a shared common
understanding, the reuse and the analysis of domain
knowledge. An ontology is, therefore, suited to
represent the BPMN metamodel.
In this context, many researchers proposed
methods for transforming a Business Process Model
(BPM) to the OWL2 ontology (Annane et al., 2019)
(BPMN-onto, 2019). These works are based on the
graphical notation (Annane et al., 2019) or on the
XPDL language (Figueiredo and Oliveira, 2018).
Although these studies are elaborated for
transforming BPMN model into ontology, they have
not fully proposed the transformation rules. In
addition, these transformation rules neglect the
semantic information related to BPMN elements. A
lack of information may reduce the number of
possible components that can be found. For instance,
the relation semantics between classes and their type
such as “is composed of” and “is a part of”, etc.
More specifically, we propose seven
transformation rules to transform BPMN model to an
OWL2 graphical representation. These
transformation rules use an annotated BPMN model
and the proposed business context that describe
semantic information related to BPMN elements.
The remainder of this paper is structured as
follows: Section 2 overviews the business context and
discusses related work. Section 3 shows the
transformation rules to generate an OWL2 graphical
representation from an annotated BPMN model with
its business context. Section 4 evaluates the quality of
the generated OWL2 graphical through the recall and
precision rates and illustrates our transformation rules
through a case study. Finally, Section 5 summarizes
the presented work and outlines its extensions.
380
Kchaou, M., Khlif, W., Gargouri, F. and Mahfoudh, M.
Transformation of BPMN Model into an OWL2 Ontology.
DOI: 10.5220/0010479603800388
In Proceedings of the 16th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2021), pages 380-388
ISBN: 978-989-758-508-1
Copyright
c
2021 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
2 BACKGROUND
2.1 BPMN Language
BPMN (ISO/IEC 19510. 2013), adopted by the OMG
group, is the most used notation for modelling
business processes (BP). The graphical objects are
organized into several categories: Activity, Data,
connecting objects (sequence/message flows),
participants (lane, pool). An activity can be a simple
representing Task or composed representing a sub-
process. In BPMN 2.0, there is different Task types
such as service Task which is used when an external
service is called to perform a task. Send task is
designed to send a message to an activity, process, or
lane, while receive task is designed to wait for a
message from an activity, process, or lane. Activities
and processes often need data objects and data store
in order to be realized. Connecting Objects (sequence
flow, message flow) connect the Flow Objects to each
other or other information to create the basic structure
of a BP. Participants represent Pools and Lanes
elements. A pool can be a specific entity or a role. It
is divided into one or more lanes.
2.2 Business Context
Before introducing the business context, we extend
the BPMN source meta-model presented in (Khlif et
al., 2018) (See Figure 1).
Figure 1: BPMN meta-model.
For each BPMN element, (Khlif et al., 2018)
associate a Description that adds a specific
information to BPMN elements in terms of the
relationships between them. The ExtendedAttributes
class specifies the properties of each BPMN element.
In (Khlif et al., 2018), the authors describe the
business context to annotate different BPMN
elements. The business context add semantic and
structural information specific to all BPMN elements.
Activity node can be simple, representing a task,
or composed that expressing a sub-process. We
enhance each activity with a business context that
contains the following information: 1) the unique
activity identifier (ID), 2) Lane ID which is the
unique identifier of the lane containing the activity, 3)
Performer (actor) ID that express the unique identifier
of the actor responsible of performing the activity, 4)
Upstream and downstream ID is the unique identifier
of the activity on which this activity directly depends,
5) extended attributes which can be a pure value or a
complex one representing a business entity, 6)
activity description indicating the relationships
between the business entities and/or the activity’s
extended complex attributes, 7) resources expressing
the data objects/stores that are required by an activity
to fulfil its goal. The resources are described in terms
of name, extended attributes and description.
The data objects/stores’ extended attributes and
description have the same semantic than the activity’s
extended attributes and description.
The lane and pool elements are described with the
following informations: 1) Unique identifier of lane
(IDL)/pool (IDP), 2) their labels, 3) Lane Description
(LD)/Pool Description (PD) to indicates the semantic
relation between the lane/pool and 4) the tasks/data
object or stores (respectively the lanes or tasks/data
object or stores) that belong to it, 5) Extended
Attributes to describe the lane/pool properties. As the
same of the extended attributes related to the activity,
each one can be a pure value or complex. The
annotated BPMN elements will be transformed into
OWL2 components.
2.3 Related Work
Many researchers proposed a number of methods for
transforming a Business Process Model to the OWL2
ontology (Annane et al., 2019) (BPMN-onto, 2019).
In this context, (Annane et al., 2019) developed
the BBO (BPMN 2.0 Based Ontology) ontology for
business process representation, by reusing existing
ontologies and meta-models like BPMN 2.0. Another
ontology (BPMN-onto, 2019) has been automatically
extracted from BPMN 2.0, but there is no
documentation about how it was generated.
Moreover, this ontology contains no annotations and
less information than the specification document.
(Figueiredo and Oliveira, 2018) propose a systematic
process for the automatic generation of an ontology
from a BP model transformed to the XML Process
Definition Language.
(Ternai, et al., 2016) propose an approach to
transform the BP into process ontology and to
combine it with the knowledge base as a domain
ontology in a well-controlled solution.
Overall, the above works focus on transforming
Transformation of BPMN Model into an OWL2 Ontology
381
BPMN model into ontology from its graphical
notation (Annane et al., 2019) or from the XPDL
language (Figueiredo and Oliveira, 2018). However,
they have not fully proposed the transformation rules.
In addition, these works rely only on the BPMN
elements to produce the corresponding component in
OWL2. In fact, several semantic relations are not
extracted such as “is composed of”,is a part of, etc.
In this paper, we propose a set of transformation
rules to transform BPMN model to an OWL2
graphical representation. These transformation rules
use business context that describe semantic
information related to BPMN elements.
3 TRANSFORMATION OF BPMN
MODEL INTO OWL2
In this section, we propose a set of rules for
transforming BPMN into OWL2. The proposed rules
exploit the business concept that describes semantic
information related to BPMN elements i.e. activity
description, lane description, etc. (Khlif et al., 2018).
Before introducing our transformation rules, we
define the following notation: in the BPMN model,
the business object is Bo
i
(Bo
j
), the extended attribute
is Exatt, and the relationship between business
objects (Bo
i
, Boj, etc) is expressed by the description
field (FD). In OWL2, the class is C(Bo
i
), data type
property is Exatt and object property is Bo
i
RBo
j
(R is
the relationship between two classes Bo
i
and Bo
j
).
The OWL2 (W3C, 2005) graphical representation
is described by WebVOWL editor, an open source
software for the visualization of ontologies.
The transformation rules are based on an
annotated BPMN model to generate an aligned
OWL2. It supposes that:
a. The description field of BPMN element follows
this linguistic pattern: «BusinessObject
+VerbalGroup + [Quantifiers]
+BusinessObject».
b. The BPMN tasks are labeled according to the
following linguistic syntax patterns:
ActionVerb +BusinessObject |NominalGroup ]
CommunicationVerb + BusinessObject
|NominalGroup + [[to ReceiverName] | [from
SenderName]]
We mean by BusinessObject any entity that
describes the business logic. The NominalGroup is a
set of pre/post-modifiers, which are centred around a
HeadWord that constitutes the BusinessObject. The
pre-modifiers (respectively post-modifiers) can be a
noun, adjective, or an ed/ing-participle (respectively,
a noun, adjective, or adverb). The VerbalGroup
indicates the relationship type between
BusinessObjects. Relationships’ semantic (semantic
of VerbalGroup) must follow these linguistic
patterns: BusinessObject+VerbalGroup+ Quantifiers
+BusinessObject. The verbal group indicates the
relation type as follows:
“is entirely made of” or “is part of” expresses
an aggregation relationship between the
business objects.
“is composed of” designates a composition
relation
“Is a” denotes the generalization/ specialization
relation. We note that the generalization can be
disjoint or complete and disjoint. If the verbal
group doesn’t belong to this set of keywords or
any synonyms, then it specifies a relation
between the business objects. The Quantifiers
gives an idea of the multiplicity.
R1. For each description field of BPMN element,
extract the object property in OWL2 and the
cardinality between the generated classes in OWL2
according to the semantic of VerbalGroup. If it is:
a. “Boi” is entirely made of “Boj” or “Boi” is part
of “Boj” or any synonyms:
We note that “is entirely made of” indicates that
the existence of the part is dependent on the whole. It
represents the relationship between the two business
objects in which the business object of an entity
consists of some objects of the other, but does not
exist in its interior. It is a combination of asymmetric
business objects and this relation indicate that a
business object is not associated with itself.
Therefore, the existence dependency leads to
transform “is entirely made of” into a pair of inverse
object properties with corresponding constraint. The
transformation rule is presented as follows (Figure 2):
Add two classes C(Bo
i
) and C(Bo
j
).
Add the object property Bo
j
RBo
i
for
representing the relationship between classes
C(Bo
i
) and C(Bo
j
), with domain is C(Bo
j
),
range is C(Bo
i
);
Set InverseFunctionalObjectProperty for
object properties Bo
j
RBo
i
.
Set IrreflexiveObjectProperty for the object
attribute Bo
j
RBo
i
.
Set AsymmetricObjectProperty for the object
attribute Bo
j
RBo
i
Add min/max cardinality constraint to the
corresponding classes.
ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering
382
Figure 2: R1 illustration: Case (a).
b. “Bo
i
is composed of Bo
j
” or any synonyms:
The expression “is composed of” in the
description field represents the relationship between
two business objects in which the business object of
an entity consists of all objects of the other business
object. It correspond to a pair of inverse object
properties in OWL2 with corresponding constraint.
The term “is composed of” is a combination of
irreflexive, so when transforming into OWL2, it must
use the “IrreflexiveObjectProperty” syntax. “is
composed of” expression is not recursive, a business
object isn’t composed with itself, so must set the
AsymmetricObjectProperty syntax. Therefore, we
propose the transformation rules as follow (Figure 3):
Add two classes C(Bo
i
) and C(Bo
j
).
Add two object properties for representing the
relationship between classes C(Bo
i
) and
C(Bo
j
): Bo
i
RBo
j
with domain is C(Bo
i
), range
is C(Bo
j
); Bo
j
RBo
i
with domain is C(Bo
j
),
range is C(Bo
i
).
Set IrreflexiveObjectProperty for the object
attribute Bo
j
RBo
i
.
Set AsymmetricObjectProperty for the object
attribute Bo
j
RBo
i
.
Add min/max cardinality constraint to the
corresponding classes.
In Figure 3, the description field contains a
relationship is composed ofbetween the “book” and
the “chapter”. Each book is composed of chapters.
These business objects are transformed to their
corresponding classes which are related by two object
properties: “is composed of” and “is founded in”. In
particular, the object property “is founded in” has the
Irreflexive and Asymmetric characteristics.
Figure 3: R1 illustration: Case (b).
c. “Bo
i
Is a/an Bo
j
”,
The expression “is a/an” in the description field
shows the top and bottom views of business objects
hierarchy. It is classified in two kinds of constraints
which are also annotated in the description field:
completeness (“disjoint and complete”) and
disjointness (“disjoint”). The top and bottom views of
business objects correspond to the class and subclass
in OWL2. The constraints “disjoint” corresponds to
the syntax “owl:disjointWith” and the constraint
“disjoint and complete”, corresponds to
“owl:disjointUnion” in OWL2. Therefore, we present
the transformation rules as follow (See Figure 4):
Add two classes C(Bo
i
) and C(Bo
j
) which is
subclass of C(Bo
i
).
If the description field contains the term
“disjoint”, use the syntax “owl:disjointWith”.
If the description field contains the term
“disjoint and complete”, use syntax
“owl:disjointUnion”..
Figure 4: R1 illustration: Case (c).
d. Else, transform the expression in the
description field into an object property.
Therefore, we present the transformation rules
as follow (See Figure 5):
Add two class C(Bo
i
) and class C(Bo
j
)
Add two inverse object properties Bo
i
RBo
j
and
Bo
j
RBo
i
which show relationship between
class C(Bo
i
) and C(Bo
j
).
Add min/max cardinality constraint to the
corresponding object properties.
In Figure 5, the description field of the task
“Create customer account” indicates a relationship
between two business objects “Customer” and
“Account”. The business object Customer has (1..n)
accounts but the business object Account is related to
one customer. These business objects are transformed
to classes in OWL2 that are related by two inverse
object properties “Has” and “concerns”.
Figure 5: R1 illustration: Case (d).
Transformation of BPMN Model into an OWL2 Ontology
383
R2. For each extended attribute of the BPMN
element, transform it to (See Figure 6):
a. Data type property in the OWL2 if the type of
the extended attribute merely represents a pure
value or primitive data type.
b. Else, a new class with the name
extendedAttributeLabel, and an object property
between the two generated classes by applying
R1 if the type of the extended attribute is
complex representing a business object.
Figure 6: R2 illustration.
R3: Transform a business object pool/lane
representing a process respectively to a class and a
subclass in OWL2.
R3.1: The business object pool/lane is transformed
respectively to class/subclass in OWL2. The class
name depends on the participant type which is a
performer or an entity. If the participant is a
performer, then the class name corresponds to the
business object name and the word “Space” or “area”.
Else, the class name is a concatenation of the business
object name and the word “Management” (Figure 7).
Figure 7: R3.1 illustration.
R3.2: The business object (pool/lane) has as many
extended attributes. It is transformed to class. The
class name corresponds to the business object
(pool/lane) name. The extended attributes are
transformed to data type properties (See R2). Each
business object (pool/lane) has a description field
which is transformed to object properties (See R1).
In Figure 8, the business object “Department”,
“Manager” and “Agent” are transformed into classes.
The description field of business object Department
(pool), defined in its business context, indicates that
the “department contains many mangers and agents”.
So that, this description is transformed into object
properties between the whole side (Department class)
and the part side (Manager and Agent classes) as well
as the minimum cardinality is 1 on the part side.
Figure 8: R3.2 illustration.
R4. For each service task, we apply R1 and R2
In addition, if the service task label respects the
renaming pattern:
R4.1: « Action verb + BusinessObject », then
transform (See Figure 9):
a. The BusinessObject into a class having the
same name of the BusinessObject.
b. The actor that performs the tasks into a class
with the same name of that actor.
c. The Action verb into an object property
between the generated classes in (a) and (b).
In figure 9, the description field of the task “Create
account” indicates that the “Agent” represents the
actor who creates the account. The Business Object
“Account” corresponds to a class with a name
“Account” in OWL2. The actor “Agent” corresponds
to a class having the same name in OWL2. The
Action verb “Create” is transformed to an object
property between the “Account” and “Agent” classes.
Figure 9: R4.1 illustration.
R4.2: « Action verb + NominalGroup», apply R1.
Then apply R4.1 on the HeadWord and transform:
a. The HeadWord into a class in OWL2.
ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering
384
b. The pre/postmodifier into data type property in
the class corresponding to the HeadWord if it
is a noun that simply represents a pure value.
The data type property has the same name of
pre/postmodifier (See Figure 10);
c. The pre/post-modifier into a class with the
name pre/post-modifier if it is a complex noun.
The relation between the HeadWord and the
pre/post-modifier to two object properties
between the two generated classes (HeadWord
and pre/postmodifier) (See Figure 11).
Figure 10: R4.2 illustration.
We note that adjectives, and ed/ing-participles
pre-modifiers as well as adjectives, and adverbs
postmodifiers are ignored.
Figure 11: R4.2 illustration.
R5. For each send/receive task, we apply R1 and R2.
In addition, when the task name follows this pattern:
R5.1: «CommunicationVerb+ BusinessObject + [[to
ReceiverName] | [from SenderName]] », transform
(See Figure 12):
a. The BusinessObject into a class with the same
name of BusinessObject.
b. The “to ReceiverName” and “from
SenderName” into two classes’ SenderName
and ReceiverName. These classes are related
to the created Business Object class in (a);
Then, add new data type property email or
phoneNumber in each class with a name
SenderName and ReceiverName;
c. The CommunicationVerb into object property
between the SenderName and ReceiverName
classes.
Figure 12: R5.1 illustration.
R5.2: « CommunicationVerb+ NominalGroup + [[to
ReceiverName] | [from SenderName]]», apply
R5.1 on the HeadWord and transform:
a. The HeadWord into a class in OWL2.
b. The pre/postmodifier into data type property in
the class corresponding to the HeadWord if it
is a noun that simply represents a pure value.
c. The pre/post-modifier into a class with the
name pre/post-modifier if it is a complex noun.
The relation between the HeadWord and the
pre/post-modifier into two object properties
between the two generated classes (HeadWord
and pre/postmodifier).
We note when this expression [[to ReceiverName]
| [from SenderName]] is omitted, then we can extract
this semantic information from the description field
of the activity element according to R1.
R6. Transform to a class each data store/object,
identified by a name, if it is not already generated.
The class name has the same data object name.
R7. If the business object (pool/lane) sends or
receives respectively a message/sequence flow
to/from another one, then transform the business
object pool/lane to class/subclass in OWL2 (See
R3.1) and the message/sequence flow into
unidirectional object property with the name
"Depends" between the associated sub classes.
Figure 13: R7 illustration.
Transformation of BPMN Model into an OWL2 Ontology
385
4 EVALUATION OF OWL2
GRAPHICAL
REPRESENTATION
In order to validate the transformation rules, we
measure the quality of the generated OWL2 graphical
representation through the recall and precision rates.
We recall that the precision is the ratio of real
elements generated by our transformation that were
identified by the expert. It indicates how accurate the
transformation rules are in the generation of OWL2
graphical representation (class, data type property and
object property) (see Formula 1).
The recall is the ratio that indicates the capacity of
our transformations to return all elements specified by
the expert. High scores for both ratios show that the
transformations return both an accurate OWL2
graphical representation (high precision), and the
majority of all relevant accurate OWL2 graphical
representation elements (high recall). It means that
the generated OWL2 graphical representation covers
the whole domain precisely in accordance to the
experts’ perspective (see Formula 2).
To have the harmonic mean of recall and
precision, we have used the F-measure. F-measure
has a parameter that sets the trade-off between recall
and precision. The standard F-measure is F1, which
gives equal importance to recall and precision (see
Formula 3). We calculate these rates according the
following equations:
FP
T
P
TP
ecision
Pr
(1
)
Re
TP
call
TP FN
(2
)
2*Re *Pr
1
Re Pr
call ecision
F
call ecision
(3
)
Where:
True positive (TP) is the number of existing
real elements generated by our transformation;
False Positive (FP) is the number of not
existing real elements generated by our
transformation;
False Negative (FN) is the number of existing
real elements not generated by our
transformation.
4.2 Case Study
To illustrate the transformation rules, we use the
BPMN modelPurchase department process
example as illustrated in Figure 14.
First, by applying R3.1, the Purchase Department
pool and the Supplier pool are transformed to classes
named “Purchase department management”, and
“Supplier Space”. Each lane (Agent, Manager) in the
Purchase Department pool is transformed to
subclasses named “Agent space” and “Manger space”
of the “Purchase department management” class.
Second, we generate four classes by applying R3.2,
which are “Purchase Department”, “Agent”
“Manager”, and “Supplier”.
The “Purchase Department” and “Agent” are
subclasses of “Agent space”. While the “Manager”
and the supplier represent respectively subclasses of
“Manager Space” and “Supplier Space” classes.
In addition, the description field of business object
“Purchase Department”, defined in its business
context, indicates that the “Purchase Department
contains many mangers and agents”. This description
is transformed into an object property
betweenPurchase Department andManager and
“Agent” classes. Furthermore, the rule R3.2 calls R2,
which adds the data type property to all classes based
on the business context of the corresponding pool and
lanes. For example, we added the data type properties
Figure 14: Purchase order Business Process model in BPMN (ISO/IEC 19510, 2013).
ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering
386
id, name and job to “Manager” and “Agent” classes,
id and designation to “Purchase Department” class.
Third, we apply R4.1 on the following service
tasks: “Create purchase order”, “Approve purchase
order”, “Deliver purchase order”, “Review
purchase order”, “Process purchase order”,
“Process payment”, “Notify payment”, “Request
quotations”, and “Select supplier”. This rule
generates three subclasses: “Purchase order”,
“Payment”, and “Quotation”. The “Supplier” class
is already generated by R3.1. Each action verb of
service tasks presented above is transformed into an
object property between the corresponding business
object and performer classes. For instance: “Create”
and “Process” represent an object property between
the “Purchase order” and “Agent” classes;
“Approve” denotes an object property between the
“Purchase order” and “Manager” classes; “Deliver”
and “Review” express an object property between
the “Purchase order” and “Supplier” classes;
“Process” and “Notify” denote an object property
between the “Payment” and “Agent” classes;
“Request” express an object property between the
“Quotation” and “Agent” classes; “Select”
represent an object property between the “Supplier”
and “Agent” classes.
Afterward, by applying R5.1, the tasks “Send
purchase order” and “Send invoice” generate an
object property with the name “Send” between the
“Agent” and “Supplier” classes.
In addition, we apply R5.1 on receive tasks
which are “Receive invoice” “Receive purchase
order request”, “Receive payment”, “Receive item”,
and “Receive quotation”. For instance: “Receive
invoice”, “Receive payment” and “Receive item”
express object properties with the name “Receive”
between the “Supplier” and “Agent” classes;
“Receive purchase order request” represent an
object property with the name “Receive” between
the “Agent” and “Customer” classes;
By applying R6, the transformation of all data
objects do not add new classes. However, R6
enhances the existing classes by calling R1 and R2,
which add data type properties, classes and object
properties. For example, we added the data type
properties deliveryDate, orderDate, orderNumber to
“Purchase order” class since these extended
attributes are pure values. Furthermore, the
extended attribute “OrderLine” is a complex entity.
According to R2, we extract a new class
“OrderLine”, and an object property between the
latter and “Purchase order”.
Figure 15 shows the generated OWL2 graphical
representation To evaluate the quality of the
generated OWL2 graphical representation, we
calculate the recall and precision rates presented in
section 4.1.
Precision=81/(81+2)=0.97
Recall =81/(81+5) =0.94
F1= (2*0.97*0.94) /(0.97+0.94)= 0.95
Figure 15: The generated OWL2 graphical representation for the purchase order process.
Transformation of BPMN Model into an OWL2 Ontology
387
Figure 16: The elaborated OWL2 graphical representation by the expert.
The high values for both ratios mean that the
generated OWL2 graphical representation covers the
whole domain precisely in accordance with the
experts’ perspective (See Figure 16). We can deduce
that the performance of our transformations
approaches the human performance.
Our generated OWL2 graphical representation
contains two object properties between the “Item”
and Purchase order request” classes. However, the
OWL2 graphical representation generated by the
expert contains: “Request Line” class, two object
properties between “Request Line” and “Item”
classes and two object properties between “Request
Line” and “Purchase order request” classes.
5 CONCLUSION
This paper focuses on transforming the BPMN model
into graphical representation of OWL2 and its
generated code. Compared to existing works, we
proposed a set of transformation rules that consider
the semantic aspects of the business process model.
To do so, we are based on the business process
context expressing the semantics relation and type.
Our future work focuses on two main axes: 1)
enhancing the transformations in order to cover use
case and sequence diagrams. 2) developing a tool for
the automatic generation of an ontology from a
business process model.
REFERENCES
Annane, A., Aussenac-Gilles, N., Kamel, M., 2019. BBO:
BPMN 2.0 Based Ontology for Business Process
Representation. In ECKM’19, 20th European
Conference on Knowledge Management. pp. 49-59.
BPMN-onto, 2019. Available at https://dkm.fbk.eu/bpmn-
ontology.
Figueiredo, L.R., Oliveira, H.C., 2018. Automatic
Generation of Ontologies from Business Process
Models. In Inter. Conf. on Ent. Information Systems.
ISO/IEC 19510. 2013. Information technology -- Object
Management Group Bus. Proc. Model and Notation.
Khlif, W., Elleuch, N., Alotabi, E., Ben-Abdallah, H.,
2018.Designing BP-IS Aligned Models: An MDA-
based Transformation Methodology. In 13th Inter.
Conf. on Evaluation of Novel Approaches to Software
Engineering, pp. 258-266.
Noy, N., McGuinness, D., 2001. Ontology Development
101: A Guide to Creating Your First Ontology.
Ternai, K., Török, M. and Varga, K., 2016. Corporate
Semantic Business Process Management, Corporate
Knowledge Discovery and Organizational Learning:
The Role, Importance, and Application of Semantic
Business Process Management, pp. 33- 57.
W3C, 2005. A survey of RDF/Topic Maps Interoperabil
ity
Proposals W3C Working Draft.
ENASE 2021 - 16th International Conference on Evaluation of Novel Approaches to Software Engineering
388