XMAP: A novel structural approach for alignment of OWL-Full ontologies

Automatic correspondence generation between two ontologies, is of great difficulty due, on one hand, to conceptual and habit differences between ontology development communities. On the other hand, alignment difficulties grew exponentially with the number and volume of the involved data. This work presents a proposition of an alignment algorithm; where the approach originality consists in taking into account the context of the alignment in order to overcome the problem of large size ontologies containing similar classes. Adding to that, we present an automatic approach to learn how to combine the linguistic and structural affinity. The weighted, sum in addition to the sigmoid function, which has to be shifted according to the weight of the linguistic affinity and to fit our input range of [0 to 1] is also used: Finally, the proposed algorithm is implemented as a Protege plug-in, and applied to align ontologies describing a steam turbine. Results are analyzed compared to manual alignment in terms of performances and pre-alignment efforts.