One of the key enabling factors of the Semantic Web vision is represented by ontologies and related metadata (ontology mappings, in particular). By means of ontologies, the computers do not really understand meanings, but they are allowed to manipulate intended semantics in an unambiguous (that is, logical) and consistent way. In order to be compatible with a deeply decentralized system as the Web, the semantic extension enabled by means of ontologies needs to be decentralized as well. The problem of (semi-)automatically computing mappings between independently developed ontologies is usually referred to as the ontology matching problem.
Due to the precise logical semantics associated with ontologies and their mappings, ontological change management cannot operate at the syntactical level only, ignoring logical consequences. Debugging techniques for unintended consequences arising in the context of ontology matching need thus to be considered as fundamental for change management. In addition, ontologies, as any other conceptual schema, are constantly evolving, and due to their complexity, they usually benefit from an incremental development, blending together ontology data and metadata change management, ontology development and debugging techniques.
Specific attention is devoted to “conservativity principle violations” detection and repair techniques for ontology-to-ontology alignments.