Databases Reference
In-Depth Information
Fig. 2.3
Screenshot of P
ROMPT
plugin while matching two university ontologies
contains several different matching algorithms in its library and the library is exten-
sible. It also assumes interaction with a user: as a user approves of certain matches,
COMA
uses this information to make further suggestions.
P
ROMPT
[
Noy and Musen 2003
](seeFig.
2.3
) is a plugin for the popular
ontology editor Protege.
2
The plugin supports tasks for managing multiple ontolo-
gies including ontology differencing, extraction, merging, and matching. P
ROMPT
begins the matching procedure by allowing the user to specify a source and tar-
get ontology. It then computes an initial set of candidate correspondences based
largely on lexical similarity between the ontologies. The user then works with this
list of correspondences to verify the recommendations or to create correspondences
that the algorithm missed. Once a user has verified a correspondence, P
ROMPT
's
algorithm uses this information to perform structural analysis based on the graph
structure of the ontologies. This analysis usually results in further correspondence
suggestions. This process is repeated until the user determines that the matching
is complete. P
ROMPT
saves verified correspondences as instances in a
matching
ontology
[
Crubezy and Musen 2003
]. The matching ontology provides a framework
for expressing transformation rules for ontology matchings. The transformation rule
support depends on the matching plugin and ontology used. In the default matching
plugin, the matching ontology simply describes the source and target correspon-
dence components and metadata, such as the date, who created the correspondence,
and a user-defined comment.
Like COMA
CC
,P
ROMPT
is extensible via its own plugin framework [
Falconer
et al.
2006
]. However, while COMA
CC
CC
supported extensibility only at the
2
http://protege.stanford.edu
.
