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

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

and Ptolemy

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

This work was funded by the European Union and the

french DGCIS through the ARTEMIS Joint Undertaking in-

side the CESAR project

https://cesarproject.eu/

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

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

 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

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-

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), Ω 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

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.

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 Ω 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