A FLEXIBLE SUPPORT OF NON CANONICAL CONCEPTS IN
ONTOLOGY-BASED DATABASES
Youness Bazhar
1
, Yamine A¨ıt-Ameur
2
, St´ephane Jean
1
and Micka¨el Baron
1
1
LIAS - ISAE ENSMA and University of Poitiers, Futuroscope, France
2
IRIT - ENSEEIHT, Toulouse, France
Keywords:
Ontology, Database, Ontology-based Database, OWL, Behavioral, Structural and Descriptive Semantics,
Canonical and Non Canonical Concepts.
Abstract:
Several ontologies are currently defined for various domains and for a wide range of applications. As a
consequence, two problems shall be addressed (1) the size of data described by ontologies is huge and (2)
existing ontologies are not all similar: they can be defined with different formalisms leading to different
types of ontologies (canonical, non canonical and linguistic ontologies). The first problem has been solved
by using databases to store data and ontologies that define the semantics of data. However, these databases,
called Ontology-Based Databases (OBDB), do not handle the second problem i.e., that different types of
ontologies exist. Indeed, they usually support a given ontology formalism without providing any mean to
handle ontologies defined in an other formalism. In this paper, we propose an extension of the OntoDB OBDB
to support the introduction of new operators that could be exploited by its exploitation language OntoQL.
Such operators can be implemented by an external program located outside the database, or could invoke a
web service. We show that these operators can be used to introduce new ontology constructors in order to
support a given ontology model inside OntoDB. As a running example, we consider the support of the OWL
non canonical constructors which are not supported in the native OntoDB OBDB.
1 INTRODUCTION
Since its introduction, the notion of ontologies has
been widely used and has regained a lot of interests
with the recent development of the Semantic Web. In-
deed, ontologies are defined in a lot of domains such
as engineering, medicine, biology or chemistry and
set up for a wide range of applications like natural
language processing, information retrieval, electronic
commerce, software component specification or in-
formation systems integration etc. The intensive use
of ontologies in a large number of domains leads to
two main difficulties. (1) The amount of data de-
scribed by ontologies can be huge, especially in do-
mains like as E-commerce, engineering or Semantic
Web. (2) Different ontology models exist to define
different types of ontologies. For example, a lot of
ontologies in engineering are defined with the PLIB
ontology model (Pierra, 2007), (Pierra and Sardet,
2010) whereas most ontologies in the Semantic Web
are defined with models such as RDF-Schema (Brick-
ley and Guha, 2004) or OWL (Dean and Schreiber,
2004). The first difficulty has been overcome by the
introduction of a new type of databases that store both
data and the ontologies that define the meaning of
these data. A lot of such databases, called Ontology-
Based Databases (OBDB), have been proposed in the
literature (e.g., Oracle (Chong et al., 2005), RStar
(Lu et al., 2007)). If these OBDBs provide scalable
storage systems for ontologies and its associated in-
stances, they are usually defined for a given ontol-
ogy model without any extension capability. Thus
they are not able to address the second issue i.e., the
wide diversity of existing ontologies. Our claim is
that solutions can be provided to handle this prob-
lem in a flexible manner if meta-modeling capabilities
are offered by the OBDB. In this paper, we consider
the OntoDB OBDB (Dehainsala et al., 2007). This
OBDB has originally been defined for the PLIB on-
tology model and thus it supports basic ontology con-
structors of classes and properties. However it does
not support the definition of non canonical concepts
(NCC) (derived or defined concepts) i.e. concepts de-
fined by a complete axiomatic definition expressed in
393
Bazhar Y., Aït-Ameur Y., Jean S. and Baron M..
A FLEXIBLE SUPPORT OF NON CANONICAL CONCEPTS IN ONTOLOGY-BASED DATABASES.
DOI: 10.5220/0003940803930398
In Proceedings of the 8th International Conference on Web Information Systems and Technologies (WEBIST-2012), pages 393-398
ISBN: 978-989-8565-08-2
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
terms of other concepts (Gruber, 1995). This kind of
concepts are particularly important in the OWL on-
tology model. For example, non canonical classes
can be defined as a union of other classes or as a
restriction on a property value. In this context, our
proposition consists in extending OntoDB to support
the introduction of new operators that could be ex-
ploited in order to introduce non canonical construc-
tors. These operators are flexible since they can be
implemented by an external program located outside
the database, or could invoke a web service. As a
proof of concepts, we show how the non canonical
constructors of OWL can be defined and implemented
on OntoDB. The remainder of this paper is organized
as follows. Section 2 reviews some related work to
our issue. Section 3 presents the OntoDB OBDB
together with its exploitation language, OntoQL, on
which our approach is based. Then, Section 4 exposes
how we have extended the OntoDB/OntoQL system
in order to support the non canonical constructors of
OWL. Section 5 presents the implementation of our
approach on OntoDB. Finally, Section 6 concludes
the paper.
2 RELATED WORK
In the last years, many OBDB architectures have been
proposed to store ontologies and instance data. We
present these architectures according to the number
of schemes used to store all information and describe
the eventual extension capabilities provided.
Type 1 OBDBs are used to store RDF data that may
contain ontology descriptions together with their
instances. Main RDF-OBDBs are Oracle (Chong
et al., 2005) or Sward (Petrini and Risch, 2007).
These OBDB follows the simple model of RDF
to store data. Indeed, a single schema composed
of a unique triple table
(subject, predicate,
object)
is used to store both ontology descrip-
tions and instance data. As RDF data may include
RDFS or OWL ontology descriptions, most of
these OBDBs provide a support for the semantics
of RDFS or OWL. This semantics is usually hard
coded using deductive rules (Chong et al., 2005)
or with external reasoners (Harris and Gibbins, 2003).
Type 2 OBDBs store separately ontology descrip-
tions and instance data in two different schemes.
Main examples are RStar (Lu et al., 2007) or On-
toMS (Park et al., 2007). The schema for ontology
descriptions depends upon the ontology model used
to represent ontologies (e.g., RDFS, OWL, PLIB).
It is composed of tables used to store each ontol-
ogy modeling primitive such as classes, properties
and subsumption relationships. For instance data,
different schemes have been proposed. They have
different scalability characteristics. These OBDBs
mainly support the usual subsumption semantics
as specified in the RDFS semantics (Hayes, 2004)
(i.e,
subClassOf
and
instanceOf
relationships).
They use different database mechanisms like views
(Pan and Heflin, 2003), labeling schemes (Park
et al., 2007) or the subtable relationships issued from
object-relational databases (Alexaki et al., 2001;
Broekstra et al., 2002). Some OBDBs address more
complex reasoning using logic-based engines (e.g.
Datalog engine) of deductive databases or OWL
reasoners (Mei et al., 2006; Volz et al., 2005; Borgida
and Brachman, 1993; Pan and Heflin, 2003).
Type 3 OBDBs. OntoDB (Dehainsala et al., 2007)
proposes to add another schema to type 2 OBDBs.
This schema called meta-schema records the ontology
model. For the ontology schema, the meta-schema
plays a role similar to the one played by the system
catalog in traditional databases. If this part may allow
the support of evolution of the used ontology, we
show in next section that the behavioral semantics of
the added constructors can not be defined.
Synthesis. As we have seen in this section, studies on
OBDBs have been mainly focused on the scalability
of these new types of databases. Considering support
of ontology, each OBDB supports the semantics of a
given ontology model using hard coded techniques ei-
ther by using database mechanisms or by relying on
an external logical engine. The aim of our work is to
provide a more flexible approach that can be followed
to support the semantics of several ontology models
in different ways. As stated previously, OntoDB pro-
vides an interesting part, the meta-schema, to support
the evolution of the used ontology model. Thus, our
approach suggests to exploit this schema for encoding
our proposed extensions. We present this architecture
in detail in the next section and show its current limi-
tations.
3 THE OntoDB/OntoQL SYSTEM
In next subsections we first present the OntoDB (De-
hainsala et al., 2007) architecture, that supports the
storage of different ontology models, ontologies and
their instances in the same database, together with
its associated language OntoQL (Jean et al., 2006).
Then, we show through a motivating example the lim-
its of this OBDB architecture and of the OntoQL lan-
guage to support the definition of the behavioral se-
mantics of OWL ontologies NCCs.
WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies
394
3.1 The OntoDB Architecture
OntoDB is an operational OBDB whose architec-
ture is divided into 4 parts according to figure 1.
The Meta-base and Instances parts are the traditional
parts available in all database management systems
(DBMS). The first one, namely the system catalog,
contains the system tables used to manage all the data
contained in the database. The second one contains
the data that represents ontologies instances. The
Meta-schema part holds ontology models (e.g. OWL
meta-schema), and, finally, the Ontologies part con-
tains ontologies conforming to the ontology models
supported by OntoDB. In OntoDB, the whole data
issued from different levels of information (meta-
schema, ontologies and instances) is stored in rela-
tional tables thanks to the services offered by the
Meta-schema part. As an example, let’s consider Fig-
ure 1. In this example, three modeling levels are rep-
resented. The meta-schema level defines two entities:
Class
and
Property
.
Class
is the abstract descrip-
tion of one or many similar objects. Classes are or-
ganized in a hierarchy linked by an inheritance rela-
tionship (
superClass
). (
Property
) describes prop-
erties of a class. Ontologies level defines the
Person
,
Student
and
Professor
classes with various prop-
erties as instances of the meta-schema previously de-
fined. Finally, the instances level defines instances
of the
Student
and
Professor
classes defined in
the ontologies level. Figure 2 shows how the data
Class
name : STRING
Property
name : STRING
superClass
properties
*
1
1
*
Student
studentNumber : STRING
Student
studentNumber : STRING
Professor
professorNumber : STRING
Professor
professorNumber : STRING
Person
name : STRING
age : INTEGER
sex : STRING
Person
name : STRING
age : INTEGER
sex : STRING
Student1 : Student
name = ‘toto’
age = 20
Sex = ‘M’
studentNumber = ‘ST567’
Professor1 : Professor
name = ‘tata’
age = 30
Sex = ‘F’
professorNumber = ‘PR345’
Student1 : Student
name = ‘toto’
age = 20
Sex = ‘M’
studentNumber = ‘ST567’
Professor1 : Professor
name = ‘tata’
age = 30
Sex = ‘F’
professorNumber = ‘PR345’
Meta-schema
Ontologies
Instances
Meta-
schema
Ontologies Instances
Meta-base
Meta-
schema
Ontologies Instances
Meta-base
Figure 1: The storage of ontology models, ontologies and
instances in OntoDB.
of the previous example are represented in OntoDB.
In this simplified example, data related to the meta-
schema part are stored in
Entity
and
Attribute
ta-
bles. Thus, the first table contains two rows for the
two entities
Class
and
Property
, and the second ta-
ble contains attributes of the defined entities. For each
entity, a corresponding table exists at the ontology
level. These tables contain instances of the different
entities of the meta-schema part. In our example, it
contains classes and properties. Finally, for each class
a corresponding table is defined at the instance level
to store its individuals.
3.2 The OntoQL Language
Since the whole data of OntoDB is stored in a
database, one can think that SQL could be used to
manipulate them. In this case, users need to have
a deep knowledge of the internal tables representa-
tion used by OntoDB. To overcome this limitation,
OntoDB has been equipped with an exploitation lan-
guage named OntoQL that hides the internal repre-
sentations and directly manipulates ontology models
and ontologies concepts. This language can be used
to create new ontologymodels and instanciate them in
order to create ontologies of different levels of infor-
mation presented in Figure 1. The following OntoQL
statements illustrate how the language proceeds:
CREATE ENTITY #Class (#name STRING,#supCls REF(#Class));
CREATE ENTITY #Property (#name STRING,#cls REF(#Class));
CREATE #Class Person (name STRING, age INT, sex STRING);
CREATE #Class Student UNDER Person (stdtNumb STRING);
CREATE #Class Professor UNDER Person (ProfNumb STRING);
INSERT INTO Professor VALUES ('tata', 30, 'F', 'PR345');
INSERT INTO Student VALUES ('toto', 20, 'M', 'ST567');
The first two statements define the meta-schema. The
entity
Class
is defined with a
name
and the class it
extends, the
supCls
. The entity
Property
has also
a
name
and is linked to the class it describes (
cls
).
In these statements, the names of entities and descrip-
tions of entities are prefixed by the character
#
. In-
deed, the meta-schema part of OntoDB can be ex-
tended and modified and thus this level of informa-
tion cannot be encoded as keywords of the OntoQL
language. The next three statements define the differ-
ent classes of our example with their properties using
the
CREATE #Class
clause. A class is defined as a
subclass of an other one using the keyword
UNDER
. Fi-
nally, the last two statements define instances of our
classes using an
INSERT INTO
syntax close to the one
of SQL.
3.3 Encoding OWL Ontologies
Canonical Concepts with OntoQL
As we have previously outlined, the aim of this pa-
per is to show how OWL can be supported by the
OntoDB/OntoQL system. In particular, we would
like to show how this system can manage the be-
havior of NCCs that are derived from other con-
cepts (e.g.
UnionClass
,
IntersectionClass
, etc.).
This subsection shows how the OWL canonical con-
cepts of OWL are represented with OntoDB/OntoQL.
AFLEXIBLESUPPORTOFNONCANONICALCONCEPTSINONTOLOGY-BASEDDATABASES
395
Entity
oid name
1 Class
2 Property
Entity
oid name
1 Class
2 Property
Attribute
oid name
3 name
4 superClass
5 name
6 itsClass
Attribute
oid name
3 name
4 superClass
5 name
6 itsClass
Class
oid name superClass
7 Person
8 Student 7
9 Professor 7
Class
oid name superClass
7 Person
8 Student 7
9 Professor 7
Property
oid name itsClass
10 name 7
11 age 7
12 sex 7
13 studentNumber 8
14 professorNumber 9
Property
oid name itsClass
10 name 7
11 age 7
12 sex 7
13 studentNumber 8
14 professorNumber 9
Student
oid name age sex studentNumber
16 toto 20 M ST567
Student
oid name age sex studentNumber
16 toto 20 M ST567
Professor
oid name age sex studentNumber
15 tata 30 F PR345
Professor
oid name age sex studentNumber
15 tata 30 F PR345
Person
oid name age sex
15 tata 30 F
16 toto 20 M
Person
oid name age sex
15 tata 30 F
16 toto 20 M
Meta-schema
Ontologies
Instances
Figure 2: Meta-schema, ontologies and instances data rep-
resentation in OntoDB.
IntersectionClass
OWLProperty
uri : STRING
OWLProperty
uri : STRING
OWLClass
uri : STRING
OWLClass
uri : STRING
domain
unionOf
intersectionOf
superClasses
subClasses
UnionClass
*
*
*
*
*
*
*
*
Figure 3: A part of a simplified OWL meta-model.
We take a part of a simplified OWL meta-model
which is presented in the figure 3. This meta-
model contains two canonical concepts:
OWLClass
and
OWLProperty
. The other concepts of this meta-
model are non-canonical. Indeed, they are derived
concepts and have to be computed from other con-
cepts. For instance, a
UnionClass
is an OWL class
made from the union of a set of
OWLClass
. Thus, the
UnionClass
is a NCC. At this level, we are capable
to encode the statements that support such concepts
description. The statements to create and to store
OWLClass
and
OWLProperty
concepts are:
CREATE ENTITY #OWLClass (#uri STRING, #superClasses REF
(#OWLClass) ARRAY, #subClasses REF (#OWLClass) ARRAY);
CREATE ENTITY #OWLProperty (#uri STRING, #domain REF
(#OWLClass) ARRAY);
The
OWLClass
defines the concept of OWL class. It
has an
uri
, a set of super classes (
superClasses
) and
a set of subclasses (
subClasses
). The
OWLProperty
defines the concept of OWL property that has an
uri
and a
domain
(the classes which the property belongs
to). These statements support the creation of struc-
tures (tables) that will permit the storage and the ma-
nipulation of OWL classes and properties in OntoDB
using the OntoQL exploitation language.
3.4 Limitation of OntoDB/OntoQL
As shown in subsection 3.2, OntoQL can be used to
add new entities in the meta-schema part of OntoDB.
For example, let’s consider the
UnionClass
concept.
It supports the definition of a class as an union of
other classes. Using this constructor, a class called
SchoolMember
could be defined as the union of the
Professor
and
Student
classes of our example. The
OntoQL statement to extend the meta-schema with
this new constructor is:
CREATE ENTITY #UnionClass UNDER #OWLClass (#unionOf
REF(#Class) ARRAY);
This OntoQL statement states that the structure of
this constructor (i.e, an
UnionClass
has the same
attributes of
OWLClass
and is defined by a set of
classes). However, we are not capable to state within
this statement nor with any other OntoQL statement,
that the instances of an
UnionClass
can be com-
puted as the union of the instances of the classes
used in its definition (Instances(C) = Instances(C1)
U Instances(C2) U ... U Instances(Cn)), and the
UnionClass
becomes a super class of the classes tak-
ing part of the union (SubClasses(C) = {C1, C2, ...,
Cn}). As a consequence, we can assert that On-
toQL does not support the behavioral semantics of
the OWL NCCs (
UnionClass
,
IntersectionClass
,
etc.). but, this semantics could be defined in opera-
tions such as functions, procedures defined or imple-
mented inside (stored procedure) or outside (external
program) the OntoDB database, or as web services.
The former is already available in some database sys-
tems but the later is still not available. For example,
we should be allowed to define the semantics of the
unionOf operator through a statement such as:
CREATE #UnionClass C1 AS
unionOf
(C2, C3);
This statement should be able to modify the structure
of C2 and C3 classes. Indeed, C1 becomes a super
class of C2 and C3. Moreover, the OWL unionOf op-
erator defines C1 individuals as the union of C2 and
C3 individuals. Therefore, there is a need of an oper-
ator
unionOfInstances
which computes the union
of C2 and C3 classes individuals in order to produce
individuals belonging to C1. This operator should be
used such as:
CREATE EXTENT OF C1 AS
unionOfInstances
(C2, C3);
The operators unionOf and unionOf-
Instances could be implemented by internal
or by external procedures or by web service invo-
cation. Therefore, extending OntoQL to support
web services, procedures, functions triggering and
constraints checking will make it possible to offer
the definition of the behavioral semantics of the
OWL ontologies NCCs. Offering such a capability is
detailed below.
WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies
396
4 EXTENDING OntoDB/OntoQL
WITH BEHAVIORAL
SEMANTICS
Our work focuses on extending the OntoDB platform
with operations that can be dynamically defined by
the user. These operations may be defined outside
database and implemented with a programming lan-
guage (e.g. Java). This extension is performed in two
steps.
4.1 Extending the OntoDB
Meta-schema Part
We Firstly have set up an application programming
interface (API) which allows to invoke, from the
OntoDB platform, external programs. Those pro-
grams are implemented with a programming lan-
guage. Then, we extended the OntoDB meta-schema
part with a new entity
EXTERNAL PROGRAM
that stores
all the information about the external program. In-
deed, these information are the function name, the
class name, the package name, and the location of the
program that contains the operation to invoke. The
OntoQL statement for extending the meta-schema
part with the
EXTERNAL PROGRAM
concept is:
CREATE ENTITY #EXTERNAL PROGRAM (#FCT NAME STRING,
#PACKAGE NAME STRING, #CLASS NAME STRING, #LOC STRING);
Thus, we can create external program and store data
related to their accessibility in OntoDB. For instance,
we can define an external program that implements
the OWL
unionOf
operator and store information
about this program in OntoDB as it is shown in the
following statement:
CREATE #EXTERNAL PROGRAM ('unionOf', 'fr.ensma.lisi.owl-
ncconcepts', 'NCOperators', 'D://owlncoperators.jar');
This statement stores the necessary information
about the
unionOf
function implemented with a
programming language and stored outside the On-
toDB platform. It supports the execution of the union
of OWL classes. Let’s consider another operator
unionOfInstances
that computes the individuals
belonging to the resulting class of the
unionOf
operation. Its definition is writting by:
CREATE #EXTERNAL PROGRAM ('unionOfInstances', 'fr...owl-
ncconcepts', 'NCOperators', 'D://owlncoperators.jar');
This statement stores information about the program
used to compute the individuals belonging to the re-
sulting class of the OWL union classes.
4.2 Extending the OntoQL Language
After extending OntoDB schemas, we allow OntoQL
to take into account external functions invocation in
the expression part. Thus, we are able to make ex-
ternal functions calls in OntoQL statements as it is
shown by the following statement
1
:
CREATE #UnionClass SchoolMember ('schoolmember') AS
unionOf (Professor, Student);
CREATE EXTENT OF SchoolMember AS unionOfInstances
(Professor, Student);
5 SUPPORTING OWL NON
CANONICAL CONCEPTS IN
OntoDB
This section present the implementation of the
UnionClass
concept. The implementations of the
other concepts are more or less similar to the one we
process. We describe the modification that affects
the ontologies elements and individuals. Then, we
present examples of statements to express structural,
descriptive and behavioral semantics of
UnionClass
concept.
UnionClass
represents a class obtained from the
union of a set of OWL classes. For instance, if we
consider that the class C1 is the union of two classes
C2 and C3 (C1 = C2 U C3), individuals of C1 are the
union of the individuals of C2 and C3 (Instances(C1)
= Instances(C2) U Instances (C3)), and C1 become
a super class of C2 and C3: SubClasses(C1) = {C2,
C3}. In order to show the complete process of defin-
ing a union class, we create two OWL classes (Pro-
fessor and Student) and make their union.
CREATE #OWLClass Professor('professor');
CREATE #OWLClass Student('student'));
The statement for the union class is:
CREATE #UnionClass SchoolMember('schoolmember') AS
unionOf(Professor, Student);
This statement modifies the structure of
Professor
and
Student
classes. It adds the
SchoolMember
class to their super classes and sets
SchoolMember
subclasses to
Professor
and
Student
. The state-
ment supporting the computation of the individuals
of
SchoolMember
is:
CREATE #EXTENT OF SchoolMember AS
unionOfInstances(Professor, Student);
1
In order to abbreviate, we use the classes names instead
of classes Uris in all the following statements
AFLEXIBLESUPPORTOFNONCANONICALCONCEPTSINONTOLOGY-BASEDDATABASES
397
6 CONCLUSIONS
In this paper, we have introduced our work on ex-
tending OBDBs offering meta-modeling capabilities
together with their exploitation languages in order to
support the representation of ontologies NCCs. Our
objective was to express the behavioral semantics of
these NCCs through operators dynamically defined
inside or outside the OBDB. In this direction, we
firstly raised lack of OBDB systems to express the
behavioral semantics of NCCs, then we extended the
OntoDB/OntoQL system in order to take into account
the representation of ontologies NCCs by defining ex-
ternal programs (defined outside the database). These
programs serve as operators, which can be invoked in
OntoQL statements. Indeed, we have shown how we
extended the meta-schema part of OntoDB in order to
take into account operations invocations, and we ex-
tended the OntoQL exploitation language in order to
make operations calls. Finally, as an implementation
of our approach, we gave the representation of OWL
ontology model in OntoDB and we proved the sup-
port of the OWL ontologies NCCs. As perspectives
of our work, we expect to exploit meta-modeling ca-
pabilities of OntoDB/OntoQLsystem in order to per-
form model transformations and process web services
orchestrations in database.
REFERENCES
Alexaki, S., Christophides, V., Karvounarakis, G., Plex-
ousakis, D., and Tolle, K. (2001). The ICS-FORTH
RDFSuite: Managing Voluminous RDF Description
Bases. In Proceedings of the 2nd International Work-
shop on the Semantic Web, pages 1–13.
Borgida, A. and Brachman, R. J. (1993). Loading data into
description reasoners. SIGMOD Record, 22(2):217–
226.
Brickley, D. and Guha, R. V. (2004). RDF Vocabulary De-
scription Language 1.0: RDF Schema. World Wide
Web Consortium. http://www.w3.org/TR/rdf-schema.
Broekstra, J., Kampman, A., and van Harmelen, F. (2002).
Sesame: A Generic Architecture for Storing and
Querying RDF and RDF Schema. In Horrocks, I. and
Hendler, J., editors, Proceedings of the 1st Interna-
tional Semantic Web Conference (ISWC'02), number
2342 in Lecture Notes in Computer Science, pages
54–68. Springer Verlag.
Chong, E. I., Das, S., Eadon, G., and Srinivasan, J. (2005).
An Efficient SQL-based RDF Querying Scheme. In
Proceedings of the 31st international conference on
Very Large Data Bases (VLDB'05), pages 1216–1227.
Dean, M. and Schreiber, G. (2004). OWL Web Ontology
Language Reference. World Wide Web Consortium.
http://www.w3.org/TR/owl-ref.
Dehainsala, H., Pierra, G., and Bellatreche, L. (2007).
Ontodb: An ontology-based database for data inten-
sive applications. In Proc. of the 12th Int. Conf. on
Database Systems for Advanced Applications (DAS-
FAA'07). LNCS. Springer.
Gruber, T. R. (1995). Toward principles for the design of
ontologies used for knowledge sharing. International
Journal of Human-Computer Studies (IJHCS), 43(5-
6):907–928.
Harris, S. and Gibbins, N. (2003). 3store: Efficient bulk
RDF Storage. In Proceedings of the 1st International
Workshop on Practical and Scalable Semantic Sys-
tems (PPP'03), pages 1–15.
Hayes, P. (2004). RDF Semantics. World Wide Web Con-
sortium. http://www.w3.org/TR/rdf-mt/.
Jean, S., ıt-Ameur, Y., and Pierra, G. (2006). Query-
ing Ontology Based Database Using OntoQL (an On-
tology Query Language). In Proceedings of On the
Move to Meaningful Internet Systems 2006: CoopIS,
DOA, GADA, and ODBASE, OTM Confederated In-
ternational Conferences (ODBASE'06), volume 4275
of Lecture Notes in Computer Science, pages 704–
721. Springer.
Lu, J., Ma, L., Zhang, L., Brunner, J.-S., Wang, C., Pan, Y.,
and Yu, Y. (2007). Sor: a practical system for ontology
storage, reasoning and search. pages 1402–1405.
Mei, J., Ma, L., and Pan, Y. (2006). Ontology query an-
swering on databases. In Proceedings of the 5th Inter-
national Semantic Web Conference (ISWC'06), pages
445–458.
Pan, Z. and Heflin, J. (2003). DLDB: Extending Relational
Databases to Support Semantic Web Queries. In Pro-
ceedings of the 1st International Workshop on Practi-
cal and Scalable Semantic Systems (PSSS'03), pages
109–113.
Park, M. J., Lee, J. H., Lee, C. H., Lin, J., Serres, O., and
Chung, C. W. (2007). An Efcient and Scalable Man-
agement of Ontology. In Proceedings of the 12th In-
ternational Conference on Database Systems for Ad-
vanced Applications (DASFAA'07), volume 4443 of
Lecture Notes in Computer Science, pages 975–980.
Springer.
Petrini, J. and Risch, T. (2007). SWARD: Semantic Web
Abridged Relational Databases. In Proceedings of
the 18th International Conference on Database and
Expert Systems Applications (DEXA'07), pages 455–
459.
Pierra, G. (2007). Context Representation in Domain On-
tologies and its Use for Semantic Integration of Data.
Journal Of Data Semantics (JODS), X:34–43.
Pierra, G. and Sardet, E. (2010). ISO 13584-32 Indus-
trial automation systems and integration Parts library
Part 32: Implementation resources: OntoML: Product
ontology markup language. ISO.
Volz, R., Staab, S., and Motik, B. (2005). Incrementally
Maintaining Materializations of Ontologies Stored in
Logic Databases. Journal of Data Semantics II,
3360:1–34.
WEBIST2012-8thInternationalConferenceonWebInformationSystemsandTechnologies
398