ATOM Ontology Merging
ATOM (automatic target-driven ontology merging) is a new approach for merging ontologies, especially taxonomies. ATOM has the following characteristics:
ATOM is match-based, i.e. it utilizes the results of a preceding match step to identify corresponding concepts in the input ontologies.
In contrast to most previous merge approaches, ATOM supports asymmetric or target-driven merging where we merge the source ontology into the target ontology. We thus give preference to the target and guarantee its complete preservation. We see important advantages for such an asymmetric approach, e.g. to incrementally extend mediator ontologies to incorporate additional source ontologies. For example, the catalog of a new online shop may be merged into the catalog of a price comparison portal. The preservation of the target ontology contributes to its stability which is valuable for applications.
ATOM can restrict the degree of semantic overlap in the merge result for improved understandability. This is achieved by giving preference to the target ontology when the same concepts are differently organized in the input ontologies.
ATOM has succesfully been applied to large ontologies such as product catalogs and biomedical ontologies.
ATOM can further utilize enhanced match mappings containing not only equivalence but also is-a and inverse is-a relationships between concepts of the input ontologies. The additional types of correspondences are used for a better placement of source concepts and to additionally reduce the semantic overlap in the merge result.
ATOM has been integrated within the COMA ontology/schema match system (see screenshot). It is now part of the COMA 3.0 Business Edition.