Title: Operationalizing Declarative and Procedural Knowledge: a Benchmark on Logic Programming Petri Nets (LPPNs)

Abstract: Modelling, specifying and reasoning about complex systems requires to process in an integrated fashion declarative and procedural aspects of the target domain. The paper reports on an experiment conducted with a propositional version of Logic Programming Petri Nets (LPPNs), a notation extending Petri Nets with logic programming constructs. Two semantics are presented: a denotational semantics that fully maps the notation to ASP via Event Calculus; and a hybrid operational semantics that process separately the causal mechanisms via Petri nets, and the constraints associated to objects and to events via Answer Set Programming (ASP). These two alternative specifications enable an empirical evaluation in terms of computational efficiency. Experimental results show that the hybrid semantics is more efficient w.r.t. sequences, whereas the two semantics follows the same behaviour w.r.t. branchings (although the denotational one performs better in absolute terms).