A Lightweight Approach for Model Checking Variability-Based Graph Transformations

Paper in proceeding, 2022

Graph transformation systems often contain large numbers of similar rules, leading to maintenance issues as well as performance bottlenecks during rule applications. Previous work introduced variabilitybased graph transformations as a paradigm for explicitly managing variability in rules, successfully
addressing these issues. However, no previous work investigated whether variability-based graph transformations can also lead to benefits during the automated analysis of graph transformations, particularly during model checking, in which the main performance bottleneck is the combinatorial
explosion arising during state space exploration.
In this paper, as an initial approach for model checking of variability-based graph transformations, we present an extension of an existing symbolic model checking technique. The existing technique, called Gryphon, converts the graph transformation system into a symbolic encoding and, from there, into the input format of a hardware model checker. We adapt Gryphon’s encoding to incorporate information on variability, which reduces the size and complexity of the overall encoding since it is now derived from a smaller set of rules (some of them being variability-based rules that represent several similar rules). In a preliminary evaluation, we show that our extension leads to performance benefits in a standard model checking scenario.