Heterogeneous Data Model Integration SIDOS – Architecture and Concept

Data models created to store or exchange data between systems are, by nature, heterogeneous in terms of standards and representation. In this research project, a novel approach called Semantic Markers, along with ontologies are used to integrate the heterogeneous data schema common in manufacturing systems. Once integrated, the schema can provide interfaces for global queries, synchronization and communication. Semantic markers are a concept similar to semantic networks, for the purpose of making data models more meaningful. By building a global library of namespaces which can be used to tag different entities and attributes of data schema, irrespective of design standard, the system makes schema, and hence data, integrable. Adding to markers, the proposed system uses blackboard technology and agent based programming to provide ontological matching between heterogeneous schemas. This system employs different ontology matching rules, applied in parallel and an aggregation technique to calculate the matching power. By supplementing semantic markers with ontology matching, the system can match schema with incomplete or missing semantic markers.