Evolution support in large-scale interoperable systems: a metadata driven approach

One of the fundamental difficulties associated with large-scale database interoperation is the cost and effort required for maintaining the federation in a consistent state after its launch. If only human intervention is to be assumed, the task of maintaining logical federation consistency becomes insurmountable as the scale increases. The need for research towards a federation evolution model which takes into account evolving semantic aspects is becoming severe. We discuss desirable properties of such an evolution model and present a semi-automatic framework which is able to detect, report and reconcile the effects of change at both local and federal conceptual levels in interoperating database systems. The basis for the model is the exploitation of intensional sources data and a flexible way of ontologically committing local to federal conceptual entities.