Y.V. Biletskiy (Canada), Z.V. Dudar, and O.G. Vorochek (Ukraine)
Interoperability, Ontology, Context, Mediation, Genetic Search
The described research is targeted to investigating an approach of context mediation for achieving semantic interoperability among sources and consumers of large scale semantically heterogeneous databases. The primary idea of this approach is performing data conversions by elimination of semantic conflicts that are detected through comparison of contexts of any source and consumer participating in data exchange [1,2]. This comparison uses ontologies as specifications of concepts, their properties and relationships between them in a knowledge domain. Also the use of ontologies is a way of specifying groups of knowledge for sharing and reuse among sources and consumers involved in data exchange only referring to those groups instead of computing their semantics each time context conversion is required. This work is concentrated on building and integrating source's and consumer's ontologies with the goal to build a domain of common cntology, which is a conceptualization independent on contexts of sources and consumers participating in data exchange. A multilevel ontological graph (ontograph) is used for formalizing ontologies because it fits in a multilevel knowledge model of contexts and it models not only semantic of links between objects but also quality of these links. A genetic approach is used for integrating ontologies into common ontology, particularly for searching meaningful ontological chains on the ontograph. This methods and models are implemented for meta-context mediation.
Important Links:
Go Back
