Comparison by partition—Finding Petri nets similarity on the basis of subnets

Abstract

Petri nets are used for modeling in many scientific areas. The existence of a large number of models gives a possibility for their comparison in search of desired or undesired characteristic structures. Unfortunately, there are only few comparison methods dedicated for Petri nets, and a search for new efficient approaches is still ongoing. We propose a new comparison algorithm dedicated for Petri net-based models of biological systems. It is based on a decomposition of compared networks into sets of subnets of biological significance that are then compared in search for a common structure. The presented algorithm has been implemented as a new module of a software tool named Holmes, which is available at https://www.cs.put.poznan.pl/mradom/Holmes/holmes.html.