Termination Analysis in Active Databases by Using Evolution Graphs

We introduce a method for rule termination analysis within active databases by using evolution graphs which simulating rule processing statically and considering both rule activation and deactivation. The evolution graph provides a more detailed analysis than activation and deactivation graphs. We propose a modification for a previous algorithm that can detect some cases of non-termination. This modification can detect more cases by selecting next rule for execution when more than one rule is ready to run regardless to the priority which is the main factor in the previous algorithm. A new selection algorithm can detect more cases that can be terminated that can not be detected by the previous algorithm