Towards semantic interoperability in an evolving environment

The interoperability between different business parties is a major issue in Enterprise Integration. As the world is not static and all things tend to evolve also enterprises evolve. Unfortunately, interoperable applications make the complex task of information system evolution even more complex. This has the consequence that changes are very expensive or that they are simply not realized. This results in static structures which are not fully suited to deal with the Enterprises needs. Changes happen on the syntactic and the semantic level. A change on the syntactic level can be a simple rename of a field in an exchanged document. A change on the semantic level is typically the case if the domain changes. This means that the meaning of data is changed. There are conflicting constraints for solutions supporting semantic interoperability in evolutionary enterprises: The necessity of expressive languages for formulating the changes and the consequences including the formulation of transformations on the one hand and the requirement for a very high throughput and the computation of a large number of instance documents on the other hand. We therefore introduce an architecture which is intended to combine scalability with complex reasoning. Ontological languages and tools, transformation of XML documents and the techniques for knowledge compilation are the basis for finding a viable and affordable solution.