The increasing availability of data from diverse sources presents new opportunities for inclusive and data-driven modeling, particularly in the area of human behavior analysis. Process discovery, as a data-driven technique of Process Mining (PM), can be used to extract behavioral and workflow models from event logs. However, this technique requires well-structured event logs that accurately reflect actual processes. In many cases, discovery approaches rely on single-source datasets and rarely integrate heterogeneous data and contextual information. This lack of integration limits the ability to extract deeper insights into human behavior and hinders comprehensive and interpretable modeling. To address this, we propose a methodological approach called Human Behavior Monitoring with Unified Log and Contextualized Process Discovery (HB-UniContex). HB-UniContex consolidates data from multiple sources, including smart sensing systems, multimedia data streams and human-generated data sources, into a unified event log that supports context-aware process discovery and detailed visualization. HB-UniContex enables a clear extraction and interpretation of human behavior in relation to contextual information. We demonstrate the applicability of this method through a case study in an Ambient Assisted Living (AAL) setting, analyzing human behavior models and examining the relationship between mood states and daily activities. Our findings reveal specific behavior patterns associated with different mood states, providing precise insights into how an individual’s daily mood correlates with certain daily activities.
Enterprises are increasingly challenged to make sense of proliferating digital traces in ways that are semantically coherent, organizationally actionable, and institutionally aligned. This research introduces CoEDiT–a commitment-driven enterprise digital twin–conceived as a reflexive modeling infrastructure that integrates conceptual abstraction, semantic technologies, and empirical data analytics. Grounded in Action Design Research, CoEDiT operationalizes a dialectical loop wherein social commitments serve as ontologically grounded constructs that link strategic intention to operational execution. Applied to air navigation services, the approach demonstrates how latent coordination structures can be surfaced and organizational learning sustained through the semantic reinterpretation of transactional data. Beyond its practical instantiation, CoEDiT contributes to current debates in Information Systems and Business Informatics by proposing a socio-technical modeling grammar that supports institutional sensing, strategic seizing, and organizational transformation. In doing so, it recasts enterprise modeling not as a representational end, but as a dynamic capability for reflexive adaptation in reindustrialized contexts.
Digital Twins (DTs) carry the promise of improved decision-making about, as well as monitoring and understanding of, the twinned entity. This makes them an attractive instrument to support the, often complex and multi-faceted, decision-making processes germane to urban planning. DTs require considerable technological investments, as they tend to be data-hungry and computing-intensive. Business-wise, such investments are only meaningful if they really add value to the intended decision-making processes. However, most current DT development approaches primarily focus on the technological potential of DTs within the limited scope of isolated business scenarios, and rarely address trade-offs between costs and benefits towards the business case, let alone the broader implications for IT/IS portfolio management. These broader considerations are crucial in urban planning contexts, which typically involve a broad ecosystem of parties, complex decision-making challenges, and pre-existing technological landscapes. Drawing on the discipline of Enterprise Architecture Management (EAM), this paper argues that architectural coordination of DT development initiatives would enable more effective valorisation of DTs potential, and more effective management of DT-related technology within a broader technological landscape. To this end, the paper discusses the vision for and an initial sketch of a specialisation of EAM for DT development.
Conceptual modeling, particularly with Entity-Relationship Diagrams (ERDs), is traditionally a manual and complex task in data engineering. This paper presents a method and tool for automating ERD generation from semi-structured data using Large Language Models (LLMs). The approach follows an iterative cycle, data ingestion, prompt design and execution, and results evaluation, with human-in-the-loop refinement to improve accuracy through expert feedback. Validation was performed on two datasets, assessing model quality via coverage, referential integrity, and Third Normal Form (3NF) compliance. Results showed high scores (0.90 and 0.93), confirming the tool’s effectiveness and the need for expert input to resolve complex modeling issues. This work offers an LLM-assisted solution to assist in complex data engineering tasks, while acknowledging the continued importance of expert oversight.
Modeling is a core task in enterprise systems engineering. The use of graphical modeling editors, however, remains cumbersome in general and poses a significant challenge for users with disabilities. Natural language processing (NLP) and intent recognition are at the forefront of making many technologies more accessible and intuitive by allowing users to engage using natural language. This paper presents a natural language interface (NLI) for speech-based UML model interaction that leverages state-of-the-art NLP technologies to enable speech-based modeling. We provide a workflow for the creation of NLIs for modeling editors, a proof-of-concept integration of this approach into the bigUML open-source modeling editor, and an empirical evaluation that shows promising results in intent recognition, the effectiveness of model creation, and usability. Thereby, this paper makes significant contributions towards more natural, inclusive, and accessible modeling.
Much research has been done on the comprehension and development of enterprise process models. In this paper, we follow up on work done previously by others on the effect of ambiguities on model comprehension. Ambiguities might lead to multiple alternative process interpretations by the readers of process models. In this paper, we will present research on techniques for collecting biometric data to investigate how we work with visual process models, some of which include lexical or visual ambiguities. We report an experiment with data from 26 persons as they interpret both unambiguous and ambiguous process models. The approach, which is based on techniques used in multi-modal learning analytics (MMLA), investigates how the type of ambiguity is correlated with data collected in parallel from EEG, eye-tracking and cameras (tracking facial landmarks). When working with ambiguous models, we generally find significantly higher memory load, cognitive load, convergent thinking, attention, information processing index, stress and confusion. Visual ambiguous models provide significantly higher numbers on these measures than lexically ambiguous models. In contrast, we find no difference in other cognitive, behavioural, and affective measures such as frustration and boredom. This work reports early investigations on the impact of ambiguity in process model comprehension. A deeper understanding of how individuals process ambiguous models can inform the design of more precise and effective visual representations, improving the usability of modelling and reducing misinterpretation.
Digital Business Ecosystems (DBEs) increasingly rely on the sharing of sensitive data between stakeholders to foster collaboration. However, to restrict access to this data, traditional security mechanisms are often not sufficient. This paper investigates one such case, part of the Horizon Europe MUSIC360 project, where policymakers want to know the economic value of music at the industry level. We propose a solution design approach that systematically links scenario-specific requirements to technical features of Privacy-Preserving Computation (PPC). A proof-of-concept experiment using the Prio+ protocol demonstrates the usability of our approach by showing that the selected implementation meets both the functional and security requirements.
There has been an increasing interest in leveraging the natural language processing capabilities of LLMs within business process modeling tasks. This paper adds to that growing body of knowledge by proposing an LLM-based approach that leverages the role-playing capabilities of LLMs to create data-annotated process collaboration models using the Business Process Model and Notation. The proposed approach consists of decomposing the modeling process into three main steps (process participant extraction, process model element extraction, and collaboration modeling) and results in the definition of Formalized Process Data to represent the process elements. The modeling approach is evaluated on several real-world Standard Operating Procedures to assess its viability in a real-world setting where relatively large documents need to be considered. The result showcases that the proposed approach is capable of creating sufficiently detailed data-annotated collaboration models.
The Meta Attack Language (MAL) provides a framework for creating domain-specific languages tailored to cybersecurity modeling and analysis. However, the discontinuation of graphical editors can disrupt workflows and hinder adoption, necessitating a systematic approach to finding suitable and sustainable replacements. This paper employs Design Science Research to identify, evaluate, and select a suitable graphical modeling tool for the MAL ecosystem. We elicited requirements from five stakeholders familiar with MAL and developed a comprehensive requirements specification artifact. This artifact was then demonstrated by systematically evaluating a set of 26 candidate tools, filtering them based on mandatory criteria. The evaluation led to the identification of ADOxx as the most suitable candidate, balancing MAL’s specific meta-modeling needs with crucial sustainable factors.
Industry 5.0 places human-centricity, sustainability, and resilience at the core of next-generation enterprise systems. Blockchain has been increasingly recognized as a potential enabler of this transition due to its decentralized architecture, transparency, and trust-enhancing capabilities. Yet, existing literature on enterprise systems remains fragmented in positioning blockchain’s role in enabling these foundational pillars. To address this knowledge gap, we screened and examined 11 peer-reviewed journal articles published between 2019 and 2024 using a scoping review methodology. We adopted the three pillars as our analytical lens. Results showed that blockchain, to some extent, can enable enterprise systems transitions to fulfill the core values of Industry 5.0. Two exemplar success stories, one in maritime logistics and one in healthcare, demonstrate feasibility but underscore unresolved scalability limits. Also, the results reveal issues needing further research and development, such as energy–latency trade-offs that hinder green transitions, underdeveloped governance models for human-in-the-loop decision-making, and fragile cross-chain interoperability that threatens operational resilience. Future research on the adoption of blockchain in enterprise systems must consider energy-efficient consensus protocols, standardized blockchain-IoT interfaces, and ethically grounded governance frameworks. Also, future research should include the development of ethical frameworks to ensure transparency, fairness, and explainability. Drawing on these insights, we outline a research agenda to move the discourse from conceptual promise to industrial practice.
The increasing advent of approaches that adopt Large Language Models for text-to-model generation is accompanied by a variety of experimental designs for their evaluation. Inspired by “The Prompt Report”, which provides a systematic account of prompting techniques for prompt engineering, we present a structured synthesis of experimental designs used to evaluate LLM-based text-to-model approaches. Through a systematic literature review, we compile a compendium of experimental designs, organized into six tree-structured categories. This compendium integrates both experimental design principles from traditional empirical research and challenges unique to LLM usage. For example, specific to LLMs, we identify studies that conduct an ablation study to isolate the effect of individual prompting techniques, as well as those that address cascading errors introduced by intermediate LLM outputs. The compendium is intended to inform the design of future evaluations in this emerging field, offering researchers structured support and guidance.
Social sustainability has often been overlooked in business process evaluation compared to the environmental and economic dimensions. However, the concept of social sustainability, which emphasizes fair and just practices towards stakeholders and society, is gaining prominence in business, yet its definition remains ambiguous, obstructing its integration into business process management. This paper investigates social sustainability indicators in current literature related to business processes. A systematic literature review reveals a diverse set of 254 indicators used to assess social sustainability, ranging from standard indicators to context-specific ones, and finds the most common indicators in the realm of employee well-being and development and human rights and labor practices. Our research highlights the need for further studies to address the challenges of defining and measuring social sustainability in the context of business processes.
Technology is rapidly evolving in modern society. A key solution for organizations to keep up with the development of technology and retain market share is to implement EA. Organizations use Enterprise Architecture (EA) to stay competitive by aligning IT with business goals. This study aimed to identify which EA-modeling notations are used in practice, focusing on how choices relate to industry, implementation year, and organizational size. A global survey with 40 EA professionals was conducted using a quantitative approach. The most commonly used modeling notations were Archimate, Business Process Modeling Notation (BPMN), and Unified Modeling Language (UML). Statistical analysis through Pearson’s Correlation test showed no significant correlation between modeling notation and industry, EA implementation year, or organizational size. However, it was noticeable that people within the public sector often used BPMN and Business Model Canvas (BMC). Furthermore, all organizations surveyed had implemented EA after 2000, and most were large companies with over 250 employees.
Participatory modeling is considered more effective for creating higher-quality enterprise models with broader stakeholder acceptance compared to traditional approaches. However, involving stakeholders directly requires more time and effort. To elaborate on the benefits, we conducted an experiment to compare the outcome of participatory enterprise modeling and traditional modeling, e.g., by interviewing stakeholders separately. We let groups of participants work in three different settings, varying the possibility of collaborating and working on a preliminary model. Considering the different conditions, we addressed the following research questions: 1) Do the models differ in size? 2) How well-elaborated are the models in terms of connections made between the elements? 3) Are the contributions of the various participants linked differently across the conditions? We found that collaboration slows down the process, which results in smaller models. Collaboration, however, leads to models that are better integrated. We found no evidence that visualization significantly supports the modeling.
Process Mining (PM) is a technique with several applications, including process discovery, which involves extracting process models from event logs that reflect how processes are actually executed. However, the resulting models are often difficult to interpret, not only due to the complexity and variability of real-world processes but also because the event logs lack sufficient information about the process. Context variables in event logs, such as resource or location, can be used to enrich the process by contextualization, making it more understandable. Selecting the appropriate level of contextual detail remains a significant challenge, as excessive contextualization can lead to complex models, while insufficient detail can obscure important insights into the process. This paper addresses this gap by proposing an entropy-based approach for context-aware process discovery, called Contextualization using Entropic Relevance (ContextER). The approach utilizes the Entropic Relevance (ER) metric to evaluate and propose contextualization that balances informativeness and complexity automatically. By evaluating the model ER, we can identify the models that optimize interpretability without hiding essential information. We demonstrate the effectiveness of the approach through a real-world case study, showing that it leads to more comprehensible and insightful process models.
Cybersecurity in OT environments is challenging, due to persistent threats and legacy systems using outdated protocols. Formal methods can verify protocol correctness, but their assumptions often fail to appear in system design. We propose a methodology that maps protocol specifications to architectural assets and simulates attacker behavior to test these assumptions. Applying our approach to a real OT use case shows how evolving infrastructure can break initial assumptions and create new risks. Simulations reveal attacker paths and support targeted mitigations, aligning formal verification with real-world resilience through an iterative process.
The development of effective User Interfaces (UIs) is of crucial importance in software engineering, as they facilitate seamless communication between humans and computers. A well-designed UI that is intuitive and easy to use ensures a better user experience and optimises the perception of software. However, the current landscape of UI development is characterised by a wide range of contexts, spanning diverse platforms, user profiles, environments and application domains. This diversity presents a significant challenge, as it increases the complexity and time-consuming nature of UI development. Model-Driven Engineering (MDE) has been used to accelerate the software development process, as well as Low-Code/No-Code (LCNC) tools, which allow developing software without coding, making it accessible to non-experts. To investigate the latest results in the generation of UIs for different contexts of use using MDE or LCNC tools, a systematic literature review (SLR) was conducted. This review analyses the strengths and limitations of the existing approaches and identifies research gaps that have not been fully addressed in the existing literature. These gaps may serve as a foundation for new research initiatives that advance the state of the art in UI development for different contexts of use.
The interest in digital twins (DTs) has recently increased among various industries. Consequently, various methods have been proposed to enhance their development and maintenance. The creation of DTs relies on integrating models and data from a variety of sources, including sensors, Internet of Things (IoT) devices, and other data streams. Model-driven engineering (MDE) is considered an effective approach for developing DTs. Our previous research indicates that current methods focus primarily on the software and systems perspective, often overlooking the importance of organizational integration. Additionally, we identified a growing need for approaches specifically designed to meet the needs of small and medium-sized enterprises (SMEs). In this paper, we introduce a method for MDE of DTs with the corresponding tool support meeting the specific requirements of SMEs. In three case studies from different domains, we demonstrate the applicability and confirm that domain experts are able to develop DTs with domain knowledge using our method.
Agile Software Development (ASD) has become the dominant approach for delivering value in iterative cycles, managing changing requirements, and coordinating distributed teams. Nevertheless, requirements management and technical debt persist as challenges. Conceptual models (CM) have been proposed as a means to address these issues; however, there is a paucity of empirical evidence regarding their systematic use and effectiveness in agile contexts. The present study investigates the perceptions of software practitioners regarding the use of CM in ASD. A total of 88 professionals actively working in software development, including project managers, analysts, developers and architects, participated in the study. Descriptive statistics were used to explore general trends, and Partial Least Squares Structural Equation Modeling was employed to examine relationships between constructs and to identify key predictors of intention to use conceptual models. The findings reveal a consistent and positive perception of CM among practitioners, with the highest ratings related to their usefulness for software requirements and system design. However, perceptions regarding their contribution to agile-specific tasks were less favorable. The findings of the present study underscore the notion that perceived usefulness and habit are the strongest predictors of intention to use CM. Furthermore, cultural factors appear to influence these perceptions. A comparison between Cuban and non-Cuban participants reveals significant disparities, particularly with regard to their propensity to adhere to organizational expectations. This study provides empirical insights into the role of CM in ASD and establishes a foundation for future work aimed at supporting their adoption and addressing cultural variations in perception.
