Mapping-Based Merging of Schemas - Schema Matching and Mapping - page 228

a. Execute any changes automatically determined as necessary by SMART.
b. Add any conflicts caused by the user's actions in step 3 to the Conflicts list.
c. Add to the ToDo list any other suggested operations or make new suggestions
based on linguistic or structural similarity.

5. Steps 3 and 4 are repeated until the ontologies are completely merged or aligned.

5.2 Chimæra

The Ontolingua Server [ Farquhar et al. 1996 ] is designed to make all parts of
dealing with ontologies easier; they have a large collection of tools to allow
users to create, view, manipulate, publish, and share ontologies. One of the tools
is Chimæra[ McGuinness et al. 2000 ], an environment for merging and testing
ontologies.
Their system, like SMART, is designed to help users merge their ontologies. The
difference is that where SMART concentrates on actually merging ontologies (e.g.,
automatically merging two classes with the same name), and Chimæra only points
out the areas where merging is likely to be required. Their goal was to build a tool
that “focuses the attention of the editor on particular portions of the ontology that are
semantically interconnected and in need of repair or further merging.” [ McGuinness
et al. 2000 ]
The authors identify a number of key features that a merging tool must sup-
port [ McGuinness et al. 2000 ]. They propose that a merging tool have support for:

Searching for names across multiple ontologies,

Renaming in a systematic fashion,

Merging multiple terms into a single term,

Focusing the user's attention on term merging based on term names,

Browsing classes and attributes,

Modifying subsumption relationships in classes and attributes, and

Recognizing logical inconsistencies introduced by merges and edits.

5.3 FCA Merge

FCA Merge [ Nebel 2001 ] from Stumme and Maedche merges ontologies based on a
lattice approach; they perform a match (in database terms) or an alignment (ontology
terms) automatically. The lattice describes both the structure of the merged docu-
ment and which elements in the ontology match according to the classes' semantic
content. The created lattice may contain both nodes that are labeled with more than
one class (indicating that merging may be required) and nodes with no correspond-
ing class in the original ontology (suggesting that the user may want to insert a new