Ontology Query Languages for Ontology-Based Databases - Data Warehousing Design and Advanced Engineering Applications - page 241

all instances of PostDupont thanks to a rule. The
meaning of this rule is the following. If a message
?p has been created by ?c (antecedent defined in
the first parameter) which is Dupont (filter defined
in the second parameter), then ?p is inferred to be
an instance of PostDupont (consequent defined
in the third parameter).

Create the class User of our example and add
some instances.

INSERT INTO rdf_example VALUES
('User', rdf:type, rdfs:Class)
INSERT INTO rdf_example
VALUES ('User', rdfs:subClassOf,
'Resource')
INSERT INTO rdf_example
VALUES ('first_name', rdf:type,
rdf:Property) ...
INSERT INTO rdf_example VALUES
('User1', rdf:type, User)
INSERT INTO rdf_example VALUES
('User1', first_name, 'Patrick') …

Requirement 3 (Linguistic
Exploitation)

Like SPARQL, the graph pattern of the RDF_
MATCH function can use string values suffixed
by a language code (e.g., @fr for a French string).
Thus, our sample queries can be written like in
SPARQL (see section 4).

Explanation. All information has to be inserted
as RDF triples. Thus, many INSERT statements
are necessary to insert the class User with its
properties and its instances.
As this example shows, usage of the DML
of SQL can be tedious to create ontologies and
their instances.

Requirement 4 (Ontology
Model Extensibility)

Like SPARQL, the RDF_MATCH function has
been designed for RDF data and thus, it consid-
ers OWL ontology as a set of triples that use the
OWL constructors defined in the OWL namespace.
However, contrary to SPARQL, predefined rule-
bases are provided to take into account semantics
of OWL constructors.

Requirement 8 (Queries
on Ontologies and on
Ontologies and Data)

Since, the RDF_MATCH function provides similar
graph pattern matching capabilities as SPARQL,
it can be used to combined ontology and data
querying.
Thus, the RDF_MATCH is a powerful exten-
sion of SQL to query RDF data. In the next section
we discuss the OntoQL exploitation language that
extends SQL following a different approach.

Requirements 5 and 6
(Compatibility with the Traditional
Database Architecture)

The RDF_MATCH function is directly integrated
in SQL (requirement 5). To use this function, a
table has to be created to store triples. Material-
ized views are automatically created to optimize
queries on this table. However the representa-
tion of triples in this table can not be customized
(requirement 6).

OntoQL

OntoQL (Jean, 2007; Jean et al., 2006) has been
defined specifically for OBDBs. To exploit data,
OntoQL has been defined in different layers:

Requirement 7 (Ontology & Data
Definition and Manipulation)

• data access at the logical level by compat-

ibility with SQL;
data access at the ontological level, CCO

The data manipulation language of SQL can be
used to create classes and instances. Example.

•

layer. Primitive concepts being mainly