Databases Reference
In-Depth Information
successfully participated in OAEI ontology matching contests; some systems even
implemented specific extensions to better solve certain contest tasks. Not shown in
the table is that most systems focus on 1:1 correspondences, although the elements
of a schema may participate in several such correspondences (the fact that element
Name in one schema matches to the combination of Firstname and Lastname in the
second schema can thus not directly be determined). Parallel matching (Sect.
3.2
)is
also not yet supported in the tools.
In the following, we provide some specific details for the considered prototypes.
4.2.1
Coma
CC
Coma
(
Aumueller et al. 2005
;
Do and Rahm 2007
) and its predecessor Coma
(
Do and Rahm 2002
) are generic prototypes for schema and ontology matching
developed at the University of Leipzig, Germany. They were among the first sys-
tems to successfully support the multimatcher architecture and match workflows
as introduced in Sect.
2
. Initially, the focus was on a metadata-based matching;
instance-based matching was added in 2006. Coma
CC
CC
supports the partitioning
and reuse approaches discussed in the previous section.
Coma
CC
is available for free for research purposes, and hundreds of insti-
tutes worldwide have used and evaluated the prototype. Surprisingly, the default
match workflow of Coma
(combining four metadata-based matchers) proved
to be competitive in many diverse areas, particularly for matching XML schemas
(
Algergawy et al. 2009
), web directories (
Avesani et al. 2005
), or even meta-models
derived from UML (
Kappel et al. 2007
). Coma
CC
CC
successfully participated in the
ontology matching contest OAEI 2006.
4.2.2
Falcon
Falcon-AO is an ontology matching prototype developed at the Southeast University
in Nanjing, China (
Hu et al. 2008
). As discussed in Sect.
3.1
, it supports a partition-
ing approach to reduce the search space and uses coefficients of the linguistic and
structural schema similarity to control the combination of matcher results. Instance-
based matching is not yet provided. In the OAEI contests from 2005-2007, it was
among the best performing systems.
4.2.3
Rimom
Rimom is an ontology matching prototype developed at Tsinghua University in
Beijing, China (
Li et al. 2009
). It was one of the first systems implementing a
dynamic selection of matchers, as discussed in Sect.
3.3
. The schema elements and
their instances are first linguistically matched; structural matching is only applied
if the schemas exhibit sufficient structural similarity. There are several methods for
