Information Technology Reference
In-Depth Information
can hamper successful interaction on a behavioral level for the purpose of
consumption or interaction of Web services.
Data-Level Mediation
The first mediation level addresses the first two types of heterogeneity identi-
fied above. As these are strongly interconnected and can be handled by similar
techniques, they are consolidated in
data-level mediation
[93]. This provides
a general mediation technique for Semantic Web applications.
The most common type of mismatch in the Semantic Web occurs owing
to usage of different terminologies by entities that need to interchange in-
formation. In ontology-based environments such as the Semantic Web, this
results from the use of heterogeneous ontologies as the terminological basis
for resource or information descriptions. A major advantage of ontologies is
that such mismatches can be handled on a semantic level by
ontology integra-
tion techniques
, explained below in more detail. Regarding the second type
of heterogeneity, of representation formats and transfer protocols, a suitable
method of handling is to lift the data from the syntactic to a semantic level
on the basis of ontologies, and then resolve the mismatches on this level [95].
Techniques for Data-Level Mediation
The central mediation techniques for the data level are semantically enabled
information integration techniques. Collectively referred to as ontology inte-
gration, the main techniques are ontology mapping, alignment, and merging,
which we briefly summarize in accordance with [102].
•
Ontology mapping
involves the creation of a set of rules and axioms that
precisely define how terms from one ontology relate to terms from another
ontology. These rules and axioms are expressed using a mapping language,
as in the example given below. Ontology mapping refers to mapping defi-
nitions only; none of the ontologies involved are changed.
•
Ontology alignment
has the role of bringing the ontologies involved into
mutual agreement. As in the ontology mapping technique, the ontologies
are kept separate but at least one of them has to be altered so that the
overlapping parts of the ontologies involved are “aligned” (i.e. they match).
•
Ontology merging
results in the creation of a new ontology that replaces
the original ontologies. The merging can be done either by unification
(all the terms from the ontologies involved are included, and mismatches
between the overlapping terms are resolved) or by intersection (only the
overlapping terms are included and their mismatches reconciled).
Example of Data-Level Mediation
In the context of the VTA, suppose that a client uses an ontology different
from that of the VTA Web service description. For illustrating the handling
Search WWH ::
Custom Search