Active rules termination analysis based on activation path and enhanced formula

Abstract

While active rules have been applied in many areas including active databases, XML document and Semantic Web, current methods remain largely uncertain of how to detect the termination when analysing rules behaviours. Some methods are based on a logical formula for a rule set, but only those non-updatable or finitely updatable variables can be contained by a formula. Some other methods are based on triggering and activation graphs, but they do not consider whether all rules of a triggering cycle can be infinitely executed during a single cyclic execution. Most methods cannot conclude termination if a rule set contains only those cycles that can be executed for a finite number of times. This paper presents the concepts of activation path and the execution sequence of a triggering cycle as well as the method to construct a formula that is able to include updatable variables, and then many termination cases that cannot be determined by previous methods, can now be detected.