Table of Contents
Fetching ...

Automating Execution and Verification of BPMN+DMN Business Processes

Giuseppe Della Penna, Igor Melatti

TL;DR

BDTransTest tackles the semantic verification gap in BPMN+DMN by automatically translating models into executable Java, enabling framework-free execution and inspection. It combines automated translation (BDTrans) with a testing framework (BDTest) that synthesizes input domains, runs multiple executions, and computes node/edge coverage. The approach is demonstrated on literature-based shipments and surgery cases, showing feasible generation-and-test cycles with substantial coverage and modest resource use. This work provides an open-source tool for end-to-end translation, execution, and verification of BPMN+DMN processes, advancing practical semantic verification in business process tooling.

Abstract

The increasing and widespread use of BPMN business processes, also embodying DMN tables, requires tools and methodologies to verify their correctness. However, most commonly used frameworks to build BPMN+DMN models only allow designers to detect syntactical errors, thus ignoring semantic (behavioural) faults. This forces business processes designers to manually run single executions of their BPMN+DMN processes using proprietary tools in order to detect failures. Furthermore, how proprietary tools translate a BPMN+DMN process to a computer simulation is left unspecified. In this paper, we advance this state of the art by designing a tool, named BDTransTest providing: i) a translation from a BPMN + DMN process B to a Java program P ; ii) the synthesis and execution of a testing plan for B, that may require the business designer to disambiguate some input domain; iii) the analysis of the coverage achieved by the testing plan in terms of nodes and edges of B. Finally, we provide an experimental evaluation of our methodology on BPMN+DMN processes from the literature.

Automating Execution and Verification of BPMN+DMN Business Processes

TL;DR

BDTransTest tackles the semantic verification gap in BPMN+DMN by automatically translating models into executable Java, enabling framework-free execution and inspection. It combines automated translation (BDTrans) with a testing framework (BDTest) that synthesizes input domains, runs multiple executions, and computes node/edge coverage. The approach is demonstrated on literature-based shipments and surgery cases, showing feasible generation-and-test cycles with substantial coverage and modest resource use. This work provides an open-source tool for end-to-end translation, execution, and verification of BPMN+DMN processes, advancing practical semantic verification in business process tooling.

Abstract

The increasing and widespread use of BPMN business processes, also embodying DMN tables, requires tools and methodologies to verify their correctness. However, most commonly used frameworks to build BPMN+DMN models only allow designers to detect syntactical errors, thus ignoring semantic (behavioural) faults. This forces business processes designers to manually run single executions of their BPMN+DMN processes using proprietary tools in order to detect failures. Furthermore, how proprietary tools translate a BPMN+DMN process to a computer simulation is left unspecified. In this paper, we advance this state of the art by designing a tool, named BDTransTest providing: i) a translation from a BPMN + DMN process B to a Java program P ; ii) the synthesis and execution of a testing plan for B, that may require the business designer to disambiguate some input domain; iii) the analysis of the coverage achieved by the testing plan in terms of nodes and edges of B. Finally, we provide an experimental evaluation of our methodology on BPMN+DMN processes from the literature.

Paper Structure

This paper contains 26 sections, 1 theorem, 2 equations, 33 figures, 1 table, 2 algorithms.

Key Result

Theorem 4.1

Let $({\cal B}, {\cal D})$ be a BPMN and a set of DMN tables linked to it, respectively, with variables $v_1, \ldots, v_k$. Suppose ${\cal B}$ fulfills the limitations of Section limitations:subsec and does not use neither parallel nor inclusive gateways, and let $J({\cal B}, {\cal D})$ be the resul

Figures (33)

  • Figure 1: BDTransTest: overall proposed methodology breakdown.
  • Figure 2: Shipment Business Process from LFM21.
  • Figure 3: Process graph data from the Shipment business process and corresponding graphical representation.
  • Figure 4: Fixing a multiple-output task using a gateway.
  • Figure 5: Get Length Decision Table used in the Shipment Business Process from LFM21.
  • ...and 28 more figures

Theorems & Definitions (1)

  • Theorem 4.1