Semantic change computation optimization in active databases

We extend the internal events method for change computation. This method derives a set of rules that incrementally compute the changes induced by an update, taking into account not only the update, but also the concept of event and the knowledge provided by the primary key integrity constraints. We further develop this idea by considering four additional integrity constraints: inclusion dependencies, exclusion dependencies, alternate keys and referential integrity constraints. We show that the knowledge provided by these integrity constraints allows us to save redundant checks and in some cases to remove whole rules. Thus, we obtain a significant improvement in performance. To the best of our knowledge there is no method for change computation with this capability