Active Databases

The increasing complexity of current information systems is supported by a new generation of database management systems (the object-relational DBMSs) that extend the static modeling capabilities of standard relational DBMS with the control of active and behavioural specifications. In particular, so-called active databases can be built by associating predicates, triggers (event-condition-action mechanisms) and stored procedures to flat or hierarchical table structures.
Unfortunately, these powerful features often are left aside by developers, who consider them too difficult to master, sometimes after unsuccessful experiments. Quite understandably, mixing event, logical and procedural mechanisms, spreading the integrity and behaviour control throughout a large and complex database, inevitably lead to a complex system the correctness of which cannot be proved and which often suffers from poor performance.
The proposed project intends to build a general model and a methodology that must help develop correct and efficient active databases through the emerging database management technologies. The model and the methodologies will be validated through a prototype CASE tool.