INCONSISTENCY IDENTIFICATION IN DYNAMIC
ONTOLOGIES BASED ON MODEL CHECKING
Mahdi Gueffaz, Perrine Pittet, Sylvain Rampacek, Christophe Cruz and Christophe Nicolle
LE2I Laboratory, UMR CNRS 5158, BP 47870, 21078 Dijon Cedex, France
Keywords: Model Checking, NuSMV Model Checker, Ontology Evolution, Ontology Design Pattern, Temporal Logic.
Abstract: The increasing use of ontologies and the cost of changes support the need to manage the evolution of
ontologies. A common kind of error in ontology evolution is the logical contradiction declined as
incoherences and inconsistences. In this paper, we propose a new approach to predict and identify the
incoherences and inconsistences in the evolution of ontologies based on temporal logic and ontology design
patterns. We implement the proposed approach using the NuSMV model checker. Based on these patterns,
we propose an automated process to guide and monitor the implementation of change while ensuring the
consistency of the evolved ontology.
1 INTRODUCTION
The semantic Web aims at organizing and
structuring the huge quantity of information present
on the Web. It consists of a semi-structured language
based on XML (Bray and al, 2006). The W3C
suggests the representation of the semantic Web in
layers. Each of them is built upon the previous
layers. Ontologies are the most important layer in
the success of the semantic Web; they are often used
in dynamic, multi-user and distributed environments.
The ontology construction goes through several
stages. Among them, the evolution step change
consists in turning the ontology more accurate and
appropriate to the domain. Ontology evolution is a
critical task because the new implementation can
lead to the apparition of incoherences and
inconsistences. Ontology inconsistency can occur
for several reasons such as: modeling errors when
correcting or adapting the ontology domain,
conceptualization or specification. An incoherence
corresponds to the existence of an unsatisfiable
concept in the ontology intension. An inconsistence
occurs when an individual exists for this
unsatisfiable concept in the ontology extension. In
the rest of the paper, we will use the term ontology
inconsistency to define the set of incoherences and
inconsistences occurring in the ontology. Ontology
evolution corresponds to the application of a
succession of change operations on the intension or
the extension of the ontology.
Leading the implementation of the changes while
maintaining the consistency of the ontology is a
crucial task and a huge cost in terms of time and
complexity. This task associated with ontology
versioning purposes is called ontology change
management. Furthermore, ontology change
management, if led by a human editor, needs to be
helped with an automatic or semi-automatic process.
Actually, it is illusive to believe that a human could
understand enough the entire conceptualization of
the ontology to be able to predict all the
consequences of the application of the changes and
avoid inconsistences or incoherences. In the
literature, one of the key objectives of the ontology
change management is to bring an automated
process to drive the application of a change while
ensuring the consistency of the evolved ontology
and its related versions.
In our contribution, we focus on incoherence
issues during the change management of the
ontology. We have defined a semi-automated
methodology based on Model Checking (Baier and
Katoen, 2008) helped by Ontology Design Patterns.
The combination of the two techniques provides the
inconsistent axiom succession patterns to retrieve in
the ontology graph in temporal logics.
This paper is about ontology inconsistency
identification in the evolution process. We first
explain our use of the term ontology and formally
define what we mean by ontology inconsistency,
before introducing our methodology. We propose a
418
Gueffaz M., Pittet P., Rampacek S., Cruz C. and Nicolle C..
INCONSISTENCY IDENTIFICATION IN DYNAMIC ONTOLOGIES BASED ON MODEL CHECKING.
DOI: 10.5220/0003919204180421
In Proceedings of the 8th International Conference on Web Information Systems and Technologies (WEBIST-2012), pages 418-421
ISBN: 978-989-8565-08-2
Copyright
c
 2012 SCITEPRESS (Science and Technology Publications, Lda.)

new way to predict and identify in the evolution of
the ontologies by using the model checking
technique. Model checking can handle complex
problems with large amounts of information, stored
as a graph, in order to verify critical systems. It will
be a good opportunity to use the model checking on
the ontology as it is represented by graphs. We use
the NuSMV model checker (Cimatti and al, 2000)
for this purpose.
The paper is organized as follows: the second
section presents the work done in the detection and
the resolution of ontology inconsistency. The third
section presents the model checking technique and
our contribution to identify the incoherences and
inconsistences in the evolution of the ontology. The
section four brings definitions on the ontology
inconsistency in description logics. These definitions
are used in section five, in which we apply our
approach in order to remove the inconsistency on a
formalized example. Finally, we end with a
conclusion and future works.
2 RELATED WORKS
Several works have been proposed to maintain the
evolution of ontologies coherent by trying to detect
and delete the occurred contradictions. In the
literature, maintaining consistency is based on the
principles of resolution presented in (Haase and
Stojanovic, 2005). When the axioms of change
applied lead to an inconsistent ontology, the
inconsistency is localized and the axiom having the
lowest degree of confidence is identified and
deleted.
The Pellet reasoner (Sirin, 2004) is the most used
for the analysis and detection of inconsistency.
Pellet is more or less accurate in its analysis based
on the types of inconsistency and does not always
give enough detail. Indeed, some logical
inconsistency types, especially those relating to
property, are not detected by Pellet. However,
combined with the change patterns defined in
(Djedidi, 2009), they are potentially supported. In
order to bind this lack, in (Djedidi, 2009), the
defined change management methodology called
Onto-Evoal (Ontology Evolution and Evaluation) is
based on modeling using these patterns.
In our proposal, we used a pattern-based
methodology in a different way. We use a formal
method, especially the model checking technique
using temporal logics to handle the identification of
logical contradictions from OWL DL logical
constraints patterns. There are few works using
temporal logics with ontology evolution. In (Plessers
and De Troyer, 2006), temporal logic is used to
represent ontology changes but the purpose is not
the inconsistency detection.
3 MODEL CHECKING
In this section, we firstly present an overview of the
model checking technique and the temporal logic.
Secondly, we present our approach using this
technique to identify ontological inconsistency.
Formal methods (Baier and Katoen, 2008) offer
great potential for an early inclusion of verification
in the design process, providing technical audit more
efficiently, and reduce the verification time. Formal
methods are highly recommended techniques for the
software development. We use the method based on
models that is the model checking method.
Model checking is a powerful tool for system
verification, as it can reveal errors that were not
discovered by other formal methods such as testing
or simulation. It uses the temporal logic to describe
the properties checking the system model. The
concepts of temporal logic are used for the first time
by Pnueli (Pnueli, 1977) in the specification of
formal properties are fairly easy to use. The
operators are very close in terms of natural language.
The formalization in temporal logic is simple
enough, although this apparent simplicity requires
significant expertise. The temporal logic allows
representing and reasoning about certain properties
of the system, so it is well-suited for the system
verification.
The model checking method examines all
relevant system states in order to check whether they
satisfy the desired property. The model checker
gives a counter example that indicates how the
model can violate the property. With the help of a
simulator, the user can locate the error and adapt the
model or the property to prevent the property
violation.
4 INCONSISTENCY
IDENTIFICATION APPROACH
This paper is about ontology inconsistency
identification in the ontology evolution. More
precisely, it is about the prediction and the
identification of inconsistent change succession
patterns in the evolution log into a system of
ontology change management. Several studies have
INCONSISTENCYIDENTIFICATIONINDYNAMICONTOLOGIESBASEDONMODELCHECKING
419

shown the importance of the ontology evolution and
the almost total missing of approaches to manage
these changes.
The evolution consists in creating and managing
different evolution of an ontology by treating the
incompatibilities between the instances, the
application and the ontologies that depend on them.
To manage the evolution of ontologies, change
management systems often generate an evolution log
for each evolution (Djedidi, 2009) (Pittet and al,
2011) (Rogozan, 2008) (Jaziri, 2009). This log aims
at tracking the changes made on the ontology. These
changes made with ontology constructors can be
additions or deletions of concepts, relations,
properties or individuals.
To identify the ontology inconsistency our
methodology has three phases. The first phase
consists in transforming the evolution log into the
NuSMV language (Gueffaz and al, 2011). The
NuSMV graph is composed by nodes and arcs. The
nodes represent the concepts and the arcs the
properties of the ontology. The second phase
consists in the generation of inconsistent axiom
succession patterns. We use a subtype of ODP that
we call change constraint pattern (CCP), to give the
validity constraints corresponding to the change
axioms. A first algorithm instantiates these
constraint patterns with the elements of the NuSMV
graph (concepts, properties, etc.). The second one
transforms all these instantiated constraint patterns
(ICP) into temporal logic formulas. Finally, the third
phase uses the NuSMV model checker to check if
one of these patterns can be found in the NuSMV
graph using the temporal logic formulas. The
NuSMV model checker steps chronologically
through each node of the graph to find a node
succession corresponding to one of the temporal
logic inconsistency patterns. It is important to notice
that nodes do not need to be direct successive
neighbors to correspond to a pattern; they just need
to appear chronologically in the same order.
Managing the effects of change implies not only
the consistency identification but also its
maintenance. The consistency maintenance consists
in proposing and implementing a set of additional
changes to resolve inconsistency. Once the
incoherencies in the ontology evolution identified,
there are many ways to resolve the inconsistency.
However, the resolution phase of the inconsistency
is beyond the scope of this paper and will be studied
in other works.
Figure 1 describes the different steps to identify
the inconsistency in the ontology evolution in our
approach. The ontology user modifies it, and all the
Figure 1: The inconsistency identification process .
modifications will be added in a graph represented
by a NuSMV language. We transform our patterns
into temporal logic formulas and give them the
NuSMV model checker in order to identify the
inconsistency in the NuSMV graph. If the model
checker detects the presence of an inconsistent
change succession corresponding to one of the
temporal logic formulas, the system gives to the user
the change succession which is in cause. The
ontology is several times verified by the NuSMV
model checker until there is no more inconsistence
or incoherence. If there is no inconsistency in the
ontology, the ontology developer creates and
updates the source ontology using the new version
of the ontology.
5 CONCLUSIONS & FUTURE
WORKS
The paper presented a new methodology to identify
inconsistency in the evolution of ontology
combining the model checking technique and the
ontology design patterns. We first introduced the
work done in the detection and the resolution of
ontology inconsistency. Several works identify and
eliminate inconsistency but sometimes do not
manage to detect all the inconsistences and
WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies
420

incoherences and the axioms in cause. Next, we
describe the model checking technique and our
global methodology. Our approach can predict all
the potential logical inconsistency in the ontology
before the addition of the incoherent change thanks
to change constraints patterns derived from ontology
design patterns. The inconsistent axiom succession
patterns are then checked by the NuSMV model
checker on the evolution log NuSMV graph,
containing the whole change succession of the
ontology. We also defined the ontology
inconsistency in description logics, and we apply our
approach on a simple example of incoherent
ontology. This allowed us to identify the succession
of axioms causing the inconsistency.
For future work, we are willing to apply our
approach to both logical inconsistency and structural
incoherency. We will also treat the inconsistency
resolution based on this methodology in a next
paper. In addition, we are looking forward to
defining and integrating all the satisfiability
constraints patterns of OWL DL in the
implementation of our solution. Finally, we aim at
implementing our solution on huge ontologies to
measure the scalability and optimize our approach.
REFERENCES
Baier, C., Katoen, J. ”Principles of Model Checking,” The
MIT Press, Cambridge, Massachusetts, London,
England, 2008.
Bray, T., Paoli, J., Sperberg-McQueen., C., M., Maler, E.,
Yergeau, F., Cowan, J., 2006. Extensible Markup
Language (XML) 1.1 (second edition) W3C
recommendation, http://www.w3.org/TR/2006/REC-
xml11-20060816/.
Cimatti, A., Clarke, E., Giunchiglia, F., Roveri, M., 2000.
NuSMV: a new symbolic model checker.
Djedidi, R., 2009. Approche d'évolution d'ontologie
guidée par des patrons de gestion de changement, PhD
Thesis.
Gueffaz, M., Cruz, C., Nicolle, C., 2011. RDF2NµSMV:
mapping semantic graphs to NµSMV model checker.
The Third International Conference on Advances in
Future Internet (AFIN 2011).
Haase, P., Stojanovic, L., 2005. Consistent Evolution of
OWL Ontologies. In A.Gomez-Perez, J. Euzenat
(Eds.), LNCS, vol.3532. The Semantic Web: Research
and Applications (pp. 182-197). Berlin, Germany:
Springer. doi: 10.1007/b136731
Jaziri, W., 2009. A methodology for ontology evolution
and versioning., The Third International Conference
on Advances in Semantic Processing(SEMAPRO
2009), pages 15-21, ISBN: 978-1-4244-5044-2,
Sliema, Malta.
Pittet, P., Cruz, C., Nicolle, C., 2011. Guidelines for a
Dynamic Ontology - Integrating Tools of Evolution
and Versionning in Ontology. International
Conference on Knowledge Management in
Information Systems (KMIS 2011).
Plessers, P., & De Troyer, O., 2006. Resolving
inconsistencies in evolving ontologies. In Y. Sure, &
J. Domingue (Eds.), LNCS: Vol.4011. The Semantic
Web: Research and Applications, Proceedings of the
3rd European Semantic Web Conference ESWC 2006
(pp. 200-214). Berlin, Germany: Springer.
Pnueli, A., 1977. The temporal logic of programs. In proc.
18th IEEE Symp. Foundations of Computer Science
(FOCS’77), Providence, RI, USA. pages 46-57.
Rogozan, D., 2008. Gestion de l'évolution d'une ontologie:
méthodes et outils pour un référencement sémantique
évolutif fondé sur une analyse des changements entre
versions de l'ontologie. PhD Thesis.
Sirin, E. et Parsia, B., 2004. Pellet: An owl dl reasoner. In
Haarslev, V. et Möller, R. (editors), Proceedings of the
International Workshop on Description Logics
(DL2004).
INCONSISTENCYIDENTIFICATIONINDYNAMICONTOLOGIESBASEDONMODELCHECKING
421