On the Marriage of Theory and Practice in Data-Aware Business Processes via Low-Code - Extended version

Ali Nour Eldin, Benjamin Dalmas, Walid Gaaloul · 2025

Read On the Marriage of Theory and Practice in Data-Aware Business Processes via Low-Code - Extended version by Ali Nour Eldin, Benjamin Dalmas, Walid Gaaloul (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-27229.txt.

Introduction

Ali Nour Eldin, Benjamin Dalmas, and Walid Gaaloul present BPMN-ProX, a low-code testing framework for business processes that depend on data. Business process models are widely used yet lack formal meaning, which blocks rigorous verification. Since data drive how process steps actually execute, the framework injects data handling directly into an executable form of the BPMN modeling language.

The platform couples advanced data handling with verification through state-of-the-art model checkers, giving formal guarantees while staying usable by non-technical experts. By marrying theory with practical modeling, it aims for more agile, reliable, and user-centered process management. The work shows how formal execution semantics can be packaged so domain experts and professionals can test data-aware processes together.

Catalog record

CREATORAli Nour Eldin, Benjamin Dalmas, Walid Gaaloul
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE54 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-27229

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

On the Marriage of Theory and Practice in Data-Aware Business Processes via Low-Code - Extended version Ali Nour Eldin, Benjamin Dalmas, Walid Gaaloul Abstract In recent years, there has been a growing interest in the verification of business process models. Despite their lack of formal characterization, these models are widely adopted in both industry and academia. To address this issue, formalizing the execution semantics of business process modeling languages is essential. Since data and process are two facets of the same coin, and data are critical elements in the execution of process models, this work introduces Proving an eXecutable BPMN injected with data, BPMN-ProX. BPMN-ProX is a low-code testing framework that significantly enhances the verification of data-aware BPMN. This low-code platform helps bridge the gap between non-technical experts and professionals by proposing a tool that integrates advanced data handling and employs a robust verification mechanism through state-of-the-art model checkers. This innovative approach combines theoretical verification with practical modeling, fostering more agile, reliable, and user-centric business process management. 1 Introduction With the substantial growth in the size and complexity of business processes due to technological advancements and increased regulatory requirements [ 23 ] , ensuring the correctness and compliance of these processes with various requirements has become a significant challenge [ 1 ] . Furthermore, in an environment where continuous adaptation is crucial, the importance of testing business processes to ensure their continued value and adaptability is increasingly critical [ 15 ] . A business process in an organization comprises interconnected activities arranged to achieve a specific goal. The process and data views are essential for understanding these processes. The process perspective, or control-flow perspective , outlines task sequences using BPMN [ 7 ] . The data perspective, similar to UML [ 22 ] , focuses on business and processing data. Integrating these views into a unified workflow modeling language enhances information representation, reduces implementation time, and supports decision-making and resource allocation. Organizations consistently face challenges in integrating data and processes [ 16 , 25 , 27 ] , which include multiple complexities. Firstly, the modeling challenge involves developing a robust framework to accurately represent intricate interactions and data manipulations within business processes. Secondly, verification ensures that these models comply with operational standards, some works proposed a verification frameworks for the process enriched by data. Thirdly, execution pertains to the practical application and monitoring of these models, ensuring they function effectively in real-time operational environments. An additional challenge is the selection of modeling constructs, which often differ from those provided by data and process modeling standards such as BPMN, SQL, and UML. Lastly, the automation and verification of such integrated models are complex for designers, often requiring assistance from developers or experts to successfully implement, verify, and automate a business model due to the lack of comprehensive tools that can address all aspects of the integration process. These dimensions can be recognized in their full complexity when it comes to the verification of the resulting integrated models [ 5 , 9 ] . Verification is crucial because data and process models that seem correct in isolation may produce errors once integrated [ 24 ] . The verification involves converting the utilized language into a well-established standard for data and process components. Specifically, this means that either the control flow is represented using Petri nets or other mathematical formalisms suitable for dynamic systems that are not easily comprehensible through front-end notations like BPMN, or the data manipulation aspects depend on abstract, logical operations that are not directly expressible in concrete data manipulation languages like SQL or OCL. Verification is crucial as it ensures the rigor and theoretical validity of executable models, essential for their flawless practical application. This step not only confirms that the models adhere to operational standards but also prevents costly errors and malfunctions before implementation. A last crucial issue is that the vast majority of the literature in this spectrum mainly provides foundational results that do not directly translate into effective verification tools. Therefore, it is crucial to have a tool that assists designers and non-technical experts in verifying the model before automating its execution. Such a tool can help detect errors and reduce maintenance time prior to the execution phase. Supporting these dimensions covers the first three parts of the business process life cycle —design, implementation, and execution— by ensuring the integration of data and processes into a verified executable model. This work addresses these limitations by introducing BPMN-ProX, a Low-Code Testing Framework designed to prove and verify executable BPMN models enriched with data. Since the goal of the low-code solution is to minimize manual coding and enable non-technical experts to create applications easily, this framework combines data and process modeling with theoretical verification and practical execution, ensuring the integrity and compliance of models. This approach allows non-technical experts to verify and execute BPMN with minimal assistance from developers. The framework is based on BPMN and UML standards for data modeling, with the artifacts of these languages translated into SQL standards to facilitate integration with verification techniques. BPMN-ProX extends an existing low-code framework to support graphical data representation within BPMN, which serves as the front-end modeling language based on the definitions from previous studies [ 20 ] . This paper presents how this language instantiates the data-related aspects of the abstract modeling language studied in [ 2 ] . The features of BPMN-ProX are derived from the requirements for concrete, executable, and verifiable data-aware process modeling languages distilled from the literature. Furthermore, an implementation of BPMN-ProX in real business process management systems is demonstrated, showing how existing artifacts can be converted to an SQL-based language. Additionally, a translator that converts a BPMN-ProX model into the syntax of MCMT 1 1 1 http://users.mat.unimi.it/users/ghilardi/mcmt/ based on the encoding rules defined in previous studies [ 2 ] is reported. MCMT is a state-of-the-art SMT-based model checker for infinite-state systems that can be used for verification. The remainder of this paper is organized as follows: Section 2 explores the existing literature. Section 3 presents the preliminaries upon which our paper is based. In Section 4 , we introduce our framework, BPMN-ProX. This is followed by a discussion of the the tools used for verification our model in Section 5 . Finally, Section 6 concludes the paper. 2 Requirement Analysis and Related Work The integration of data and processes is a longstanding area of research at the intersection of BPM, data management, process mining, and formal methods. Given the focus on verification and the integration of verification methods within an executable process, the discussion is confined to relevant works addressing process and data modeling, process and data verification, and process and data execution in BPMN. Table 1 outlines six key requirements (R1–R6) for modeling languages used in the verification, execution, and deployment of executable models within Business Process Management Systems (BPMSs). Requirements R1, R5, and R6 are inspired by prior works focused on the integration of process and data modeling and execution [ 21 , 20 , 13 ] , while requirements R2 to R4 are derived from foundational studies on verification frameworks [ 13 , 17 ] . These six requirements capture critical capabilities expected from modeling approaches, including support for expressive

Read the complete text