FIRST STEPS TOWARD A VERIFICATION AND VALIDATION
ONTOLOGY
Mounira Kezadri and Marc Pantel
IRIT, Universit´e de Toulouse, ENSEEIHT 2 rue C. Camichel, Toulouse, France
Keywords:
Ontology, Modeling languages, Verification and validation technologies.
Abstract:
This paper presents the key elements of an ontology that formalizes part of the knowledge about behavioural
modeling and the associated verification and validation technologies. It summarizes the objects, concepts,
and other entities that are assumed to exist in this area of interest and the relationships that hold among them.
We propose a classification of different modeling formalisms and a representation of possible verification and
validation methods. A system is represented using several views conforming to different modeling languages.
Its properties can be assessed with verification and validation technologies.
1 INTRODUCTION
Verification and validation (V&V) is the process of
checking that a system satisfies its requirements.
Verification relates a system implementation and its
specification whereas validation relates a system and
its end users. If the requirements have been specified
then the system is verified against this specification,
and the specification is validated against its end
users. This paper presents the preliminary elements
of a verification and validation Ontology (VVO
1
),
which represents a knowledge sharing (Neches
et al., 1991) initiative for the V&V domain. The
classification is mainly based on the state of the art
of CESAR
2
and Ptolemy
3
projects and others web
sources. The ontology (Guarino and Giaretta, 1995)
defines the verification and validation methods that
can be applied on a system whose behaviour has been
modeled. VVO is being used as a communication
language, as a foundation for other engineering
ontologies, and later to constitute a global verification
and validation platform for a set of V&V techniques.
In this paper we describe the conceptualization of
the ontology, its design and motivate the major
representation choices. Our contributions are: 1) the
classification of system’s description formalisms,
2) the classification of Verification and Validation
1
This work was funded by the European Union and the
french DGCIS through the ARTEMIS Joint Undertaking in-
side the CESAR project
2
https://cesarproject.eu/
3
http://ptolemy.eecs.berkeley.edu/
methods, 3) the definition of relations between V&V
technologies, formalisms and properties descrip-
tion languages. Our intention is first to make the
knowledge of the behaviour modeling and V&V
domains shareable and reusable and then to use
the
VVO
as a guideline for choosing and applying
dynamically the adapted V&V technologies in order
to ease the development of safety critical systems.
The VVO contains more than 250 classes. Is was
developed using the Prot´eg´e tool-kit. It is available
for reuse, comments and extension proposals at
http://www.irit.fr/Mounira.Kezadri/Ontologies/
VVO.owl.
The rest of this paper is organized as follows: In
Section 2 the general architecture of the VVO, the
definition of global concepts and relations between
them are given. In Section 3 we present the case study
of verification for Petri Nets. Section 4 discusses re-
lated work. Conclusion and future work appear in
Section 5.
2 THE GENERAL STRUCTURE
OF THE VVO
In this section, we present the general architecture
of the ontology. A number of ontology modeling
methods have been proposedin the literature (Gomez-
Perez et al., 1991). The Web Ontology Language
(OWL) (McGuinness et al., 2004) under the Prot´eg´e
4
4
http://protege.stanford.edu/
440
Kezadri M. and Pantel M..
FIRST STEPS TOWARD A VERIFICATION AND VALIDATION ONTOLOGY.
DOI: 10.5220/0003102904400444
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2010), pages 440-444
ISBN: 978-989-8425-29-4
Copyright
c
2010 SCITEPRESS (Science and Technology Publications, Lda.)
Figure 1: Global architecture of VVO.
4.1 Alpha is used to build the
VVO
and the Jambalaya
tab
5
from the version 3.4.3 to obtain Figure1.
The most typical kind of ontology has a taxon-
omy defining the classes, objects and relations among
them and a set of inference rules powering reasoning
functions (Lee et al., 2001). The first phase for the
ontology definition is the enumeration of important
terms and relationships of the domain (in our case the
domain of V&V). The figure 1 shows the global struc-
ture of our ontology, it represents its key concepts and
the relations between them. One system can be de-
scribed by different abstractions, every abstraction is
expressed using a view conformingto a modeling lan-
guage. We can express properties that specifies parts
of the system. The system with, or without, associ-
ated properties can then be assessed with some V&V
technique.
In the following subsections, we present the defi-
nition and the hierarchy of the principles concepts of
the VVO.
2.1 An Ontology for Formalisms
Description
System Concept. A system is a collection of com-
ponents organized to accomplish a specific func-
tion or set of functions. One system can be com-
posed from one or several components and de-
scribed using several views expressed in modeling
formalism.
Component Concept. The characteristic properties
of a component are that it is both a unit of in-
5
http://protegewiki.stanford.edu/wiki/Jambalaya 2.6.0
dependent deployment, and of third-party com-
position; and it has no (externally) observable
state (Clemens et al., 1998). A component can
be described using views expressed in different
formalisms, specified by properties and assessed
with V&V technologies.
Formalism Concept. We consider this part of
the ontology as a basis for structuring and con-
structing a domain-specific modeling tools. The
Formalism’s ontology collects a large number of
widely-used formalisms for system’s behaviour
modeling. As the number of formalisms is quite
important, we propose a classification of these
formalisms shown in Figure 2.
Figure 2: The first level of formalisms hierarchy.
We present the example of the hierarchy of
the automata formalism which is a sub-class
of automata-based formalisms, this hierarchy is
composed of: B¨uchi automata, cellular automata
(Cervelle et al., 2010), finite automata (Lawson,
FIRST STEPS TOWARD A VERIFICATION AND VALIDATION ONTOLOGY
441
Figure 3: Automata based hierarchy.
2005), hierarchical automata (Mikk et al., 1997),
Muller automaton (Perrin, 2004), stochastic au-
tomata (D’Argenio and Katoen, 2005), timed au-
tomata (Bengtsson and Yi, 2003), automata
(Krishnan et al., 1994) and hybrid automata (Alur
et al., 1993). The hierarchy is illustrated in Fig-
ure 3. Each kind of this formalisms has special
characteristics and have his own hierarchy.
2.2 An Ontology for Views
This part is mainly derived from the IEEE standard
(Hilliard, 2000).
System Abstraction Concept. Most systems are
generally very complex. Several abstractions are
usually required in order to manage their descrip-
tion. An abstraction conforms to a view and it is
modelled in some formalism.
View Concept. Is the representation of a whole sys-
tem from the perspective of a related set of con-
cerns. In our ontology a view must be conforming
to at least one view point.
ViewPoint Concept. Is a specification of the con-
ventions used for constructing and exploiting a
view. A pattern ortemplate from which to develop
individual views by establishing the purposes and
audience for a view and the techniques for its cre-
ation and analysis.
2.3 An Ontology for Verification and
Validation Techniques
V&V Concept. The most important and the largest
part of this work is the classification of the V&V
techniques. A wide variety of V&V strategies
and techniques are available. A V&V technique
(method or technology) can be applied to one or
Figure 4: V&V techniques Hierarchy.
several abstractions of a system depending on the
formalism used to describe the system and the
property that must be assessed. We mean by ver-
ification techniques the tools to verify in some
measure that a system satisfies some specification.
We mean by validation technique the mean for the
modeling language user to check that the model is
a correct rendering of the idea he wanted to ex-
press. The goal is to increase the confidence that
we haveon the developedsystem, this can be done
with different approaches. Figure 4 shows the hi-
erarchy that we propose for the V&V techniques.
Property Concept. To describe a property, we
can use several Property Description Languages
(PDL). As an example we present the part of tem-
poral logics that we use in the case study in Sec-
tion 3. We can differentiate several temporal log-
ics, the Linear Temporal Logic (LTL), Compu-
tational Tree Logic (CTL) and State Event-LTL
(SE-LTL) are subclasses of temporal logic pre-
sented in Figure 5.
A SE-LTL formula (Chaki et al., 2004) can be as-
sociated to a Petri Nets illustrated in Figure 6 to ex-
press a verification property. This kind of property
associated to Petri Nets can be verified using a model
checking technique.
3 CASE STUDY
We instantiated the V&V concepts of the ontology
with several V&V tools, like the TINA toolbox that
KEOD 2010 - International Conference on Knowledge Engineering and Ontology Development
442
Figure 5: Temporal Logics Hierarchy.
Figure 6: Petri Nets hierarchy.
we use in our case study. To elaborate this test, the
DL Query
6
tab is used. We present the case for the
verification of Petri Nets which are also known as
place/transition nets. Petri Nets is one of the for-
malisms of our ontology, it is composed of several
sub-classes: algebraic Petri nets, coloured Petri nets,
time Petri nets, timed arc Petri nets and untimed Petri
nets.
We can explore very easily our ontology, for ex-
ample, if we want to verify a SE-LTL (State/Event
LTL) formula on a time Petri nets system, our query
and the result are expressed in Figure 7. The query
states that the V&V technique must support the time
Petri nets formalism and the SE-LTL formula and that
a property on the time Petri nets formalism can be de-
fined in the SE-LTL logic as presented in Figure 7.
The result is TINA-Selt, it is the V&V tool specialised
in the verification of SE-LTL formulas on time Petri
nets systems.
6
http://protegewiki.stanford.edu/wiki/DL Query
Figure 7: A query example.
4 RELATED WORK
By domain ontology we mean ontologies that exist in
a specific domain of interest, it defines the basic terms
and relations comprising the vocabulary of a topic
area, as well as the rules for combining terms and re-
lations to define extensions to the vocabulary. For in-
stance, a number of domain ontologies are available
on the Internet. But, to our knowledge, there is no
ontology reported anywhere for the V&V domain. A
Software Testing Ontology in UML for A Software
Growth Environmentof Web-Based Applications was
proposed by Hong Zhu in (Huo et al., 2003). We have
reported the terms in relation with the concept test of
V&V techniques of the VVO presented in Figure 4.
5 CONCLUSIONS
We are working now on improving the VVO, the next
step is to validate the ontology by linking existing
with V&V tools and experimenting with various sys-
tems. We will submit the ontologyto the system mod-
eling and V&V community in order to gather as much
feedbacks as possible. The ontology includes concep-
tual foundationsfor formalisms and V&V techniques,
but is designed for the moment for knowledge sharing
purposes.
REFERENCES
Alur, R., Courcoubetis, C., Henzinger, T., and Ho, P. (1993).
Hybrid automata: An algorithmic approach to the
specification and verification of hybrid systems. Hy-
brid systems, pages 209–229.
Bengtsson, J. and Yi, W. (2003). Timed automata: Seman-
tics, algorithms and tools. Lectures on Concurrency
and Petri Nets, pages 87–124.
FIRST STEPS TOWARD A VERIFICATION AND VALIDATION ONTOLOGY
443
Cervelle, J., Formenti, E., and Guillon, P. (2010). Ul-
timate Traces of Cellular Automata. Arxiv preprint
arXiv:1001.0251.
Chaki, S., Clarke, E., Ouaknine, J., Sharygina, N., and
Sinha, N. (2004). State/event-based software model
checking. In Integrated Formal Methods, pages 128–
147. Springer.
Clemens, S., Dominik, G., and Stephan, M. (1998). Compo-
nent software: Beyond object-oriented programming.
D’Argenio, P. and Katoen, J. (2005). A theory of stochastic
systems part I: Stochastic automata. Information and
computation, 203(1):1–38.
Gomez-Perez, A., Fern´andez-L´opez, M., and Corcho, O.
(1991). Ontological engineering. AI Magazine, 36:56.
Guarino, N. and Giaretta, P. (1995). Ontologies and
knowledge bases: Towards a terminological clarifica-
tion. Towards very large knowledge bases: knowledge
building and knowledge sharing, 1(9):9.
Hilliard, R. (2000). Ieee-std-1471-2000 recommended
practice for architectural description of software-
intensive systems. IEEE, http://standards. ieee. org.
Huo, Q., Zhu, H., and Greenwood, S. (2003). A multi-agent
software environment for testing Web-based applica-
tions. COMPSAC-NEW YORK-, pages 210–215.
Krishnan, S., Puri, A., and Brayton, R. (1994). Deter-
ministic automata vis-a-vis deterministic Buchi au-
tomata. Algorithms and Computation, pages 378–386.
Lawson, M. (2005). Finite automata. Handbook of net-
worked and embedded control systems, pages 117–
143.
Lee, T., Hendler, J., Lassila, O., et al. (2001). The semantic
web. Scientific American, 284(5):34–43.
McGuinness, D., Van Harmelen, F., et al. (2004). OWL web
ontology language overview. W3C recommendation,
10:2004–03.
Mikk, E., Lakhnechi, Y., and Siegel, M. (1997). Hierar-
chical automata as model for statecharts. Advances in
Computing ScienceASIAN’97, pages 181–196.
Neches, R., Fikes, R., Finin, T., Gruber, T., Patil, R., et al.
(1991). Enabling technology for knowledge sharing.
AI magazine, 12(3):36.
Perrin, D. (2004). Infinite words: automata, semigroups,
logic and games. Academic Press.
KEOD 2010 - International Conference on Knowledge Engineering and Ontology Development
444