Verification of data-aware process models: Checking soundness of data Petri nets
Nikolai M. Suvorov,Irina A. Lomazova
DOI: https://doi.org/10.1016/j.jlamp.2024.100953
IF: 1.088
2024-04-01
Journal of Logical and Algebraic Methods in Programming
Abstract:During recent years, significant research has been done in the direction of enriching the traditional control-flow perspective of processes with additional dimensions, such as data and decisions. To represent data-aware process models, various formalisms have been proposed. In this work, we focus on Data Petri nets (DPNs), an extension to a Petri net with data. Data in a DPN is set as variable values. Process activities, represented as transitions, can inspect and update variable values. This work is dedicated to soundness verification of data-aware process models represented as DPNs. We show the flaw in one of the algorithms for checking soundness of DPNs with variable-operator-variable conditions. The algorithm fails to detect some types of livelocks and, thus, is incorrect in the general case. In this report, we propose an advanced version of this algorithm, which correctly verifies soundness of DPNs and which can also be used for DPNs has composite conditions on transitions. To verify soundness, the algorithm refines a DPN by splitting some of its transitions, constructs an abstract state space of a refined DPN, and inspects it for soundness properties. The report justifies correctness of the proposed algorithm for DPNs with variables of real data type or any finite data types. The algorithm is implemented, and the results of its performance evaluation demonstrate practical applicability of the algorithm for process models of small and medium sizes.
computer science, theory & methods,logic