An OWL Implementation of OntoUML and BPMN Models to Unify

Representation of Structure and Behavior of Complex Domains:

Application to Routing Protocols

Mohamed Bettaz

Faculty of Information Technology, Czech Technical University in Prague, Thakurova 9, Prague, Czech Republic

Keywords:

Ontology, UFO, gUFO, OntoUML, BPMN, OWL, Dynamic Routing Protocols, Turtle, SPARQL.

Abstract:

The objective of this paper is twofold. First, we propose an approach to map BMPN to OWL, and then we use

OntoUML and BPMN (in their ontological form) to demonstrate the effectiveness of our approach through its

application to a complex and irregular problem domain, namely dynamic routing protocols. This allows us to

query their structural and dynamic aspects (and reason about them) in a uniform and transparent manner. It

should be noted that for sake of readability, the models representing parts of routing protocols are intentionally

kept simple, emphasizing key concepts and their relationships.

1 INTRODUCTION

OntoUML (Barcelos et al., 2013) is a conceptual on-

tology language used for the speciﬁcation of struc-

tural aspects of problem domains. Business Process

Modeling Notation (BPMN) (Silver, 2011) is a visual

language used for modeling process aspects of prob-

lem domains. The web ontology language (OWL)

(W3C, 2025a) is a computational and knowledge rep-

resentation ontology language provided with a formal

semantics. A transformation of OntoUML into OWL

was proposed in (Barcelos et al., 2013). In this con-

tribution we propose an approach to transform BPMN

into OWL. The idea behind such a goal is that map-

ping OntoUML and BPMN into a “universal” knowl-

edge representation ontology language allows us to

build knowledge-based systems for complex domains

that are query-able and lending themselves to uniform

reasoning. Moreover, speciﬁcations from models rep-

resenting different system structures and behaviors

and which are also expressed in different languages

become interoperable. In a recent contribution (Bet-

taz and Maouche, 2025), we showed how to use an as-

sociation of OntoUML and SoaML (Service-oriented

architecture Modeling Language) to specify IoT sys-

tems. OntoUML was used for the speciﬁcation of the

structure, while SoaML was used for the speciﬁcation

of the behavior of these systems. In this work we en-

visage to use BPMN (instead of SoaML) for its read-

https://orcid.org/0000-0003-1346-0244

ability, ﬂexibility and visual appeal. The objective of

this paper is twofold. First, we propose an approach

to map BMPN to OWL, and then we use OntoUML

and BPMN (in their ontological form) to demonstrate

the effectiveness of our approach through its appli-

cation to a complex and irregular problem domain,

namely dynamic routing protocols. The OWL ontol-

ogy resulting from the models built using OntoUML

and BPMN is implemented in Turtle and queried on a

Fuseki (SPARQL) server. The implementation of this

ontology can serve as a knowledge-based system that

can be used to reason on dynamic routing protocols

(cf.section 8)

The rest of the paper is organized as follows. Sec-

tion 2 summarizes basic knowledge on BPMN, RDF

(Resource Description Framework), OntoUML and

routing protocols. This section aims at making the

paper as self-constrained as possible; parts or all of

this section can be skipped by the reader who is in-

troduced to the topics covered in this section. In sec-

tion 3, we present the used research method and for-

mulate our research hypotheses. Section 4 presents

related work. In section 5, we present and moti-

vate the approach we use to map BPMN to OWL.

Section 6 presents a case study (acting as a proof

of concept) of the approach described in section 5.

An implementation of the ontology resulting from the

models built in our case study is given in section 7.

Some concluding remarks and future work are then

outlined.

254

Bettaz, M.

An OWL Implementation of OntoUML and BPMN Models to Unify Representation of Structure and Behavior of Complex Domains: Application to Routing Protocols.

DOI: 10.5220/0013504800003970

In Proceedings of the 15th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH 2025), pages 254-261

ISBN: 978-989-758-759-7; ISSN: 2184-2841

2 BACKGROUND

2.1 BPMN

BPMN is used for the modeling of process aspects of

problem domains. It comprises the following mod-

eling constructs: activities, events and gateways. An

activity can typically be atomic, in which case it is

called a task, or decomposable, in which case it is

called a subprocess (Author, 2024). Events fall into

three main types: start, intermediary, and end. From

these three basic types we can form various sub-types.

Gateways are used to control the ﬂow in BPMN pro-

cesses, which means that they deal with branching

in the processes. In a BPMN process, we typically

encounter the gateways exclusive, inclusive, parallel,

event-based, exclusive event-based and parallel event-

based. For more details the kind reader can consult

(Silver, 2011).

2.2 Resource Description Framework

RDF (W3C, 2025b), which is the main building block

of OWL, is a data model designed by the World Wide

Web Consortium (W3C) for describing resources us-

ing semantic triples (subject, predicate, object). A re-

source is anything that can be uniquely identiﬁed by

an IRI (Internationalized Resource Identiﬁer), which

is a URL-like object identiﬁer consisting of a name

space and a local identiﬁer. RDFS (Resource Descrip-

tion Framework Schema)(W3C, 2025) adds structure

to RDF by distinguishing resource types and deﬁn-

ing their properties, thus forming a simple ontological

language including a taxonomy of individuals. OWL

is a family of languages (OWL Lite, OWL DL and

OWL Full), that allow to specify not only types and

relationships among them but also constraints such

as multiplicities and generalization sets among oth-

ers. For a description of the full set of OWL language

constructs, the reader may consult (W3C, 2025a)

2.3 OntoUML

OntoUML is a UML proﬁle using stereotypes deﬁned

by the Uniﬁed Foundational Ontology (UFO) (Guiz-

zardi et al., 2021). An OntoUML model consists of

(shared) concepts and relationships used to describe

structural aspects of problem domains. In OntoUML

both concepts (UML classes) and relationships (UML

associations) are stereotyped. In OntoUML mod-

els, stereotypes are enclosed in double angle brack-

ets. According to the stereotypes, types are either sor-

tals or non-sortals. Instances of sortal types have one

principle of identity. Instances of non-sortal types can

have different principles of identity. Sortal and non-

sortal types can be either rigid or anti-rigid. Many

notions of UFO and OntoUML are formalized. For

more details on UFO and OntoUML, the kind reader

can consult for instance (Guizzardi et al., 2021) and

related literature.

2.4 Routing Protocols

Communication protocols in general and routing pro-

tocols in particular are considered as a complex do-

main. Using more than one language to specify their

various facets is not new in the discipline. Indeed,

from the very beginning of the 1990’s, specialists of

the domain tried to associate different speciﬁcation

languages to cope with their various facets. As ex-

amples we mention Lotos (Turner, 1993), an associ-

ation of abstract data types and process calculus; Es-

telle (Turner, 1993), an association of Pascal and ﬁnite

state automata; ECATnets (Bettaz et al., 1992), an as-

sociation of rewriting logic and abstract data types.

In this paper we focus on (centralized) dynamic rout-

ing protocols, where contrary to static routing proto-

cols, the elaboration of the forwarding tables is the re-

sponsibility of the routers. The entity responsible for

elaborating these tables is called control plane. The

tables elaborated by the control plane are handed to

the data plane, which is the entity responsible for us-

ing these tables to forward data packets.

3 METHOD AND HYPOTHESIS

In this work we use a qualitative research method aim-

ing at providing BPMN with a declarative semantics

based on the use of RDF semantic triples. Details

on the used approach are given at the very beginning

of section 5. The proposed method is supported by

a case study acting as a proof of concept (Hustadt,

2024). Our ﬁrst research hypothesis can be stated as

follows. While some research works tend to deﬁne the

semantics of BPMN through its mapping to other lan-

guages (such as Petri nets, abstract state machines or

activity diagrams), we propose to use RDF, a declar-

ative language that is also used for knowledge repre-

sentation and reasoning (KR

) (Salihoglu, 2024). Our

second research hypothesis can be stated as follows.

While many research works use different languages

to specify various aspects of problem domains, we

propose to use languages that can be uniﬁed through

their mapping to a “universal” knowledge representa-

tion ontology language.

An OWL Implementation of OntoUML and BPMN Models to Unify Representation of Structure and Behavior of Complex Domains:

Application to Routing Protocols

255

4 RELATED WORK

In (de Brock, 2024), the author elaborates a declar-

ative semantics for BPMN, based on “modeling the

semantics of (individual) actions as state transitions

represented as states”.

In (Kchaou et al., 2021), the authors propose rules

to transform BPMN models into OWL2 graphical

representation. Dixit: “The transformation rules are

based on an annotated BPMN model to generate an

aligned OWL2.”

In (Such

anek and Pergl, 2021), the authors pro-

pose an approach that allows to create a BORM on-

tology by showing how to represent the knowledge

carried by its process models in RDF. For the con-

ceptualization using RDF, the work in (Kchaou et al.,

2021) and the work in (Such

anek and Pergl, 2021) are

similar as regards to the transformation into OWL.

Compared to (de Brock, 2024), (Kchaou et al.,

2021), and (Such

anek and Pergl, 2021), our transfor-

mation approach from BPMN to RDF uses a declar-

ative approach based directly on the deﬁnition of an

ontology in terms of concepts that are abstractions of

objects (activities, events and gateways), and relation-

ships that are representations of lines with arrows con-

necting these objects.

As regards to the building of ontologies for dy-

namic routing protocols, there are no contributions,

to the best of our knowledge, using associations of

OntoUML and BPMN. Same observation concerning

the uniform conceptualization of both OntoUML and

BPMN using RDF semantic triples.

5 MAPPING BPMN TO OWL

Our mapping approach is based on a simple but ef-

fective observation, that from an ontological point of

view, a BPMN model can be perceived as a set of con-

cepts and relationships, where concepts are materi-

alized by BPMN objects (i.e., activities, events, and

gateways), while relationships are formed by lines

with arrows linking these objects. Expressing our

concepts and relationships in the form of RDF triples

provides our BPMN models with declarative seman-

tics. In the following conceptualization of objects (in

their various forms) and relationships, only samples

of triples are given for illustration as suggested by

some of the reviewers. A full speciﬁcation can be pro-

vided (on demand) by the author.

5.1 Conceptualization of BPMN Tasks

and Activities

The concept task forms a type, the (high-level) in-

stances of which are themselves types. These in-

stances are called manual, user, service, and script;

their (low-level) instances (or individuals) are units of

work called activities. Indeed, the icons used to de-

note a manual, user, service or script activity, ﬁgur-

ing on the graphical representations of BPMN activ-

ities are sorts of stereotypes leveraging the semantics

of such activities. Such stereotypes (regrouping indi-

viduals of the same sort) will be considered as OWL

classes. The RDF triples will thus take the following

form. A colon preceded by a blank means that we do

not worry about the IRIs for the moment; they will

be deﬁned in section 7. Examples of description of

BMPN tasks by semantic triples are:

:manualTask rdf:type owl:Class

:activity rdf:type :manualTask

:activity rdf:type :userTask

:activity rdf:type :serviceTask

:activity rdf:type :scriptTask

:manualTask rdfs:subClassOf :task

. . .

Subprocesses are classes of classes: ad-hoc sub-

processes, loop subprocesses, multi-instance sub-

processes, compensation subprocesses, compensation

and ad-hoc subprocesses. Such stereotypes will also

be considered as OWL classes. Their instances are

called structured-activities. Our semantic triples take

the following form.

:adHocSubprocess rdf:type owl:Class

:structured-activity rdf:type :adHocSubprocess

. . .

5.2 Conceptualization of BPMN Events

BPMN start, intermediate and end events are seen as

different types of events; start event is in turn seen as

a subclass of catching event and also as a subclass of

non interrupting event; intermediate event is seen a

subclass of catching event, as a subclass of throwing

event and as a subclass of non interrupting event; end

event is seen as a subclass of throwing event.

Examples of semantic triples describing BPMN event

modeling constructs are:

:event rdf:type :startEvent

:startEvent rdfs:subClassOf :catching

. . .

SIMULTECH 2025 - 15th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

256

5.3 Conceptualization of BPMN

Gateways

The concept gateway is a type with six instances,

named exclusive, parallel, inclusive, event based, ex-

clusive event based, and parallel event based (cf. sec-

tion 2.1). Examples of semantic triples used to specify

BPMN gateways are:

:exclusive rdf:type :gateway

:parallel rdf:type :gateway

:inclusive rdf:type :gateway

. . .

5.4 Conceptualization of BPMN

Relationships

Relationships are morphisms (abstractions of BPMN

lines with arrows) the domain and range of which are

types representing concepts (abstractions of various

BPMN modeling constructs such as tasks, events, and

gateways). BPMN lines with arrows will be repre-

sented by a semantic triple as follows.

:< arrow name > rdf:type owl:ObjectProperty

rdf:domain :< type name >

rdf:range :< type name >

6 CASE STUDY: DYNAMIC

ROUTING PROTOCOLS

We begin with the speciﬁcation of structural aspects

and then proceed to the speciﬁcation of process as-

pects, focusing on key concepts and relationships.

Unless otherwise stated, we will henceforth use the

term routing protocols to refer to dynamic routing

protocols.

6.1 Speciﬁcation of the Structures

Structural aspects are speciﬁed by OntoUML mod-

els that describe key concepts of the dynamic rout-

ing protocols’ domain, such as static and dynamic

routing, autonomous systems, centralized vs decen-

tralized algorithms, and routing tables. For lack of

space we will just specify the model depicted by ﬁg-

ure 1. From the speciﬁcation of this model, it should

be clear how to specify the models depicted by the

other ﬁgures. Figure 1 should be read as follows.

IP routing protocols (<< Category >> IP Routing)

are either static (<< Kind >> Static) or dynamic

(<< Kind >> Dynamic). Static routing protocols

are characterized by the creation of routing tables

at boot time (<< Mode >> Routing Table Created

Figure 1: Static and dynamic routing.

at Boot time). Dynamic routing protocols are char-

acterized by the use of complex routing algorithms

(<< Mode >> Complex Routing Algorithms). In

OntoUML, the stereotype Mode is used with spe-

cial classes to emphasize signiﬁcant entity proper-

ties, hence the association of the stereotype Char-

acterization (<< Characterization >>) with the re-

lationship between the given entity and its empha-

sized property. Dynamic routing algorithms (<<

Mode >> Complex Routing Algorithms) are used to

compute an optimal route (<< Relator >> Optimal

Route Determinator) among the set of all possible

routes. To this end, our dynamic routing algorithms

should advertise (<< Role >> Routing Information

Adviser) and learn (<< Role >> Routing Informa-

tion Learner) routing information about IP subnets

to and from neighbouring routers. An optimal route

is determined using a given metric (<< Quality >>

Metric) (cf. (Kurose and Ross, 2021) and (Peterson

and Davie, 2021)). As for the stereotype Mode, it

is worth mentioning that classes with the stereotype

Quality are also special classes used to emphasize

signiﬁcant entity properties. The difference between

<< Mode >> and << Quality >> is that the second

is measurable, while the ﬁrst is not.

Note: The OCL (Open Constraint Language)

predicate (OMG, 2025) in the model depicted by ﬁg-

ure 4 states that a router is not a neighbour of itself.

6.2 Speciﬁcation of the Behaviors

In (Bettaz and Maouche, 2025), we utilise SoaML (a

language used for specifying cyber-physical systems)

for the speciﬁcation of process aspects of IoT sys-

An OWL Implementation of OntoUML and BPMN Models to Unify Representation of Structure and Behavior of Complex Domains:

Application to Routing Protocols

257

Figure 2: Autonomous systems (AS).

Figure 3: Centralized vs decentralized algorithms.

Figure 4: Generation of routing tables.

tems. In this contribution we utilise BPMN models

for the speciﬁcation of process aspects of routing pro-

Figure 5: Basics from routing protocols.

Figure 6: Protocol service.

tocols. These models are depicted in ﬁgures 5, 6, 7,

8, 9, and 10. Figure 5 depicts a scenario with three

participants (routers R1, R2, R3). In this scenario R2

is supposed to be the router starting the process by ad-

vertising available routing information to both R1 and

R3. R1 is supposed to be the router ending the pro-

cess after updating its routing table and advertising

updated routing information.

The model depicted in ﬁgure 6, describes the pro-

tocol service; its interpretation follows in a straight-

forward way from the the interpretation of the used

BPMN constructs and their (sequential) composition.

The chosen routing protocol scenario supposes a

network consisting of four routers (R1, R2, R3, R4)

interconnected according to the following topology.

R1 has two adjacencies: R2 and R3. R2 has three

adjacencies: R1, R3, and R4. R3 has three adjacen-

cies: R1, R3, and R4. R4 has two adjacencies: R2

and R3. This scenario describes how R4 proceeds to

build its routing table (cf. Figure 7). Once R4 re-

ceives all the LSAs from all the routers, it proceeds

by building its LSDB and its weighted graph. Then

the minimal-weight graph algorithm is executed and

the routing table created.

6.3 Conceptualization of the Structures

For illustration, we consider the OntoUML model de-

picted in ﬁgure 1. This model describes notions of

static and dynamic routing. The conceptualization

is based on gUFO, a lightweight implementation of

UFO in OWL (Almeida et al., 2020). In the following

we just write down a semantic triple representing a

class and a semantic triple representing a relationship

(cf. ﬁgure 1). From these, it should be clear how to

write down semantic triples for the other classes and

relationships.

example class) class Static

:Static rdf:type owl:Class ;

rdfs:subClassOf gufo:Object ;

SIMULTECH 2025 - 15th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

258

rdfs:subClassOf :IP Routing ;

rdf:type gufo:Kind ;

rdfs:subClassOf owl:Restriction ;

owl:onProperty :characterizes;

owl:qualiﬁedCardinality 1;

owl:onClass :Default Route.

. . .

example relationship) relationship “character-

izes” (between class Static and Class Default Route)

:charaterizes rdf:type owl:ObjectProperty ;

rdf:type gufo: Characterization ;

rdfs:domain :Static ;

rdfs:range :Default Route .

. . .

6.4 Conceptualization of the Behaviors

We consider the BPMN model of the protocol de-

picted in ﬁgure 7 and present samples illustrating the

conceptualization of few concepts (tasks/processes,

events and gateways) and relationships (lines with ar-

rows).

6.4.1 Subprocesses and Activities

We just give the semantic triples representing subpro-

cess A and activity buildLSDB. From these, it should

be clear how to write the semantic triples for the other

subprocesses and activities.

(a1) subprocess A

:A rdf:type :adHocSubprocess .

:adHocSubprocess rdf:type owl:Class ;

rdfs:subClassOf :subProcess .

. . .

(a4) activity buildLSDB

:buildLSDB rdf:type :scriptTask .

:scriptTask rdf:type owl:Class ;

rdfs:subClassOf :task .

. . .

6.4.2 Events

Figure 7 shows three events (a start event, an interme-

diate event, and an end event).

(b1) start event s1

:s1 rdf:type :startEvent .

:startEvent rdfs:subClassOf :event .

(b2) intermediate event R4gotAllTheLSAs

:R4gotAllTheLSAs rdf:type :intermediateEvent ;

rdf:type :message ;

rdfs:subClassOf :catching .

:intermediateEvent rdfs:subClassOf :event .

(b3) end event t1

Figure 7: Routing protocol model with collapsed subpro-

cesses.

Figure 8: Subprocess A - expanded.

Figure 9: Subprocess B - expanded.

:t1 rdf:type :endEvent .

:endEvent rdfs:subClassOf :event .

6.4.3 Gateways

Figure 7 shows two gateways (both of them are paral-

lel).

(c1) parallel gateway g1

:g1 rdf:type :parallelGateway .

:parallelGateway rdfs:subClassOf :gateway .

(c2) parallel gateway g2

:g2 rdf:type :parallelGateway .

:parallelGateway rdfs:subClassOf :gateway .

An OWL Implementation of OntoUML and BPMN Models to Unify Representation of Structure and Behavior of Complex Domains:

Application to Routing Protocols

259

Figure 10: Subprocess C - expanded.

6.4.4 Relationships

Figure 7 shows thirteen lines with arrows (a1, a2, ...,

a13). In the following we write down the triples for

only two of them. From these, writing the triples for

the remaining lines should be clear.

arrow a1

:(s1, g1) rdf:type owl:ObjectProperty ;

rdf:domain :startEvent ;

rdf:range : parallelGateway .

arrow a2

:(g1, A) rdf:type owl:ObjectProperty ;

rdf:domain :parallelGateway ;

rdf:range : adHocSubprocess .

. . .

7 IMPLEMENTATION

The ontology is implemented with Turtle and queried

on a Fuseki (SPARQL) server running on a DELL

computer (Intel Core i7, 16.0 GB RAM). The code

can be provided (on demand) by the author. For lack

of space, we just consider the implementation of the

behaviors.

The code is saved in a ﬁle called ﬁg7.ttl, con-

taining the description of the preﬁxes, followed by

the deﬁnition of the concepts and the relationships

according to Turtle syntax.

A sample of the implementation results is presented

by the screenshot depicted in ﬁgure 11, and a sample

of query corresponding to these results is presented

by the screenshot depicted in ﬁgure 12. Figure 11

shows the name of the server (Apache Jena Fuseki),

the name of the uploaded Turtle ﬁle (Fig7.ttl), the

size of this ﬁle (4.64kb), and the number of uploaded

triples (74) that effectively corresponds to the number

of semantic triples saved in the ﬁle. The status

(100.00 in green) states that the implementation ran

without errors. In turn ﬁgure 12 shows the preﬁxes

Figure 11: Routing protocol implementation.

Figure 12: Routing protocol implementation - sample

query.

used in the query and the query itself. Downloading

the results of the query from the server shows the

following.

processName

http://uri/subprocess#A

http://uri/subprocess#B

http://uri/subprocess#C

Indeed, the routing protocol depicted in ﬁgure 7

shows effectively that we have three ad hoc subpro-

cesses, namely A, B and C. This result can be ex-

ploited when implementing the protocol in a program-

ming language by launching the execution of the three

subprocesses in parallel.

8 CONCLUSION AND FUTURE

WORK

The approach we proposed to map BPMN to OWL

is a ﬁrst result of our contribution. Indeed, map-

ping BPMN to a computational and knowledge rep-

resentation ontology language allows to unify prob-

lem domain representations expressed in different

languages. Moreover, the implementation of ontolo-

gies resulting from such representations can serve

as a knowledge-based system to reason about dy-

namic routing protocols. Indeed, from our speciﬁca-

tions we can infer new sentences in form of triples.

For instance, from the speciﬁcation presented in sec-

tion 6.4.2, it should be clear that we can infer

the following (and many other) triple(s) stating that

R4gotAllTheLSAs is a catching event in our protocol

speciﬁcation.

SIMULTECH 2025 - 15th International Conference on Simulation and Modeling Methodologies, Technologies and Applications

260

:R4gotAllTheLSAs rdf:type :catching.

The effectiveness of our approach has been demon-

strated by its application to a complex problem do-

main, namely dynamic routing protocols. Our mod-

els were (partially) implemented in Turtle and queried

on a Fuseki (SPARQL) server. Examples of our im-

plementation and associated queries are presented as

screenshots. In addition to this, we have shown how

to use OntoUML to build ontology fragments for

the structural part of dynamic routing protocols, and

BPMN (in its OWL form) to build an ontology for

the process part of this kind of protocols. This can

be considered as a second result of our contribution.

From the problem domain perspective, our ontology

can serve as an enrichment of the domain knowledge

on which routing protocols are based. It is known

that to develop an application/system for a domain,

one must have (or collaborate with people who have)

some expertise (knowledge) of the domain.

We plan in the near future to formalize some as-

pects of what is presented in this work. We plan to

exploit two ideas. The ﬁrst idea consists in express-

ing the protocol service as a many-sorted algebra, and

expressing the implementation of the service (i.e., the

protocol itself) as a many-sorted algebra. The ob-

jective is to show that both algebras are isomorphic,

thus proving formally that the protocol implements

its service correctly. The second idea consists in us-

ing rewriting logic (Diaconescu, 2025) to show that

the semantic triples used to specify the protocol could

be derived from the semantic triples used to specify

the protocol service. This is an operational approach

permitting to show that the protocol implements cor-

rectly its service. The other beneﬁt of this idea is

to “visualize” the parallelism inherent to the proto-

col by exploiting the concurrent computing permitted

by rewriting logic. This could be efﬁciently applied

for instance to simulate the parallelism exhibited by

ﬁgures 7, 8, 9, and 10.

ACKNOWLEDGEMENTS

The author thanks the ﬁve anonymous reviewers for

their valuable comments that helped improve the ﬁnal

version of this article.

REFERENCES

Almeida, J. P. A., Guizzardi, G., de Almeida Falbo, R., and

Sales, T. P. (2020). gUFO: A Lightweight Implemen-

tation of the Uniﬁed Foundational Ontology (UFO).

https://www.researchgate.net/publication/339747587

gUFO A Lightweight Implementation

of the Uniﬁed Foundational Ontology UFO.

Author, A. (2024). BPMN Activity

Types Explained. https://www.visual-

paradigm.com/guide/bpmn/bpmn-activity-types-

explained/.

Barcelos, P. P. F., dos Santos, V. A., Silva, F. B., Monteiro,

M. E., and Garcia, A. S. (2013). An Automated Trans-

formation from OntoUML to OWL and SWRL. pages

130–141.

Bettaz, M. and Maouche, M. (2025). Towards an

Ontological-based CIM Modeling Framework for IoT

Applications. Informatica An International Journal

for Computing and Informatics, 48(4):663–684.

Bettaz, M., Maouche, M., Soualmi, M., and Boukebeche,

M. (1992). Using ECATNets for Specifying Commu-

nication Software in the OSI Framework. In 1992

Fourth International Conference on Computing and

Information, pages 410–413, USA. IEEE Computer

Society.

de Brock, B. (2024). Assigning Declarative Semantics to

Some UML Activity Diagrams and BPMN Diagrams.

In Shishkov, B., editor, Business Modeling and Soft-

ware Design, pages 65–82, Cham. Springer Nature

Switzerland.

Diaconescu, R. (2025). Institution-independent Model The-

ory. Springer, 2nd edition.

Guizzardi, G., Benevides, A., Fonseca, C., Porello, D.,

Paulo, J., Almeida, A., and Sales, T. (2021). UFO:

Uniﬁed Foundational Ontology. In Applied Ontology

1 (2021). IOS PRESS.

Hustadt, U. (2024). Research methods in com-

puter science. https://cgi.csc.liv.ac.uk/ ull-

rich/COMP516/notes/lect06.pdf.

Kchaou, M., Khlif, W., Gargouri, F., and Mahfoudh, M.

(2021). Transformation of bpmn model into an owl2

ontology. In ENASE, pages 380–388. INSTICC,

SciTePress.

Kurose, F. and Ross, K. (2021). Computer Networking - A

Top-Down Approach. Pearson, 8th edition.

OMG (2025). Object constraint language.

Peterson, L. L. and Davie, B. S. (2021). Networks - A Sys-

tems Approach. MK, 6th edition.

Salihoglu, S. (2024). In praise of RDF.

https://blog.kuzudb.com/post/in-praise-of-rdf/.

Silver, B. (2011). BPMN Method & Style. Cody-Cassidy

Press, 2nd edition.

Such

anek, M. and Pergl, R. (2021). Representing BORM

Process Models using OWL and RDF. In KEOD, vol-

ume 2, pages 170–177. SciTePress.

Turner, K. J. (1993). Using Formal Description Techniques:

An Introduction to ESTELLE, LOTOS and SDL (Wiley

Series in Communication and Distributed Systems).

Wiley, 1st edition.

W3C (2025). RDF Schema 1.1.

https://www.w3.org/TR/rdf-schema/.

W3C (2025a). OWL Web Ontology Language Reference.

https://www.w3.org/TR/owl-ref/.

W3C (2025b). RDF 1.2 Concepts and Abstract Syntax.

https://www.w3.org/TR/rdf12-concepts/.

An OWL Implementation of OntoUML and BPMN Models to Unify Representation of Structure and Behavior of Complex Domains:

Application to Routing Protocols

261