Keywords: Cyber Physical and Human Systems, OntoUML, Knowledge Engineering, Smart Grid, Smart Buildings
Astract: Energy-focused Cyber-Physical-Human Systems (CPHSs) depend on understanding how events, states, and temporal transitions shape system behaviour, particularly in settings where human actors, such as building occupants, grid operators, and domain experts, influence state evolution. However, relevant causal knowledge in such settings is often tacit, informally held by domain experts, and difficult to translate into formal conceptual models. Existing OntoUML-based methodologies provide semantic precision but offer limited structured guidance for eliciting such knowledge or for modelling evolving temporal states in domains with both physical and human-driven dynamics. To address this gap, we propose Tacit Knowledge Tree Onto (TKTOnto), a structured four-stage knowledge engineering workflow that integrates shared conceptualisation, object and event identification, and explicit temporal state modelling using established OntoUML constructs such as Phase and Situation. TKTOnto does not introduce new OntoUML primitives; rather, it provides systematic procedures for eliciting tacit, expert-based causal explanations through semi-structured interviews and translating them into OntoUML models grounded in Unified Foundational Ontology (UFO) semantics. The workflow is applied to two real-world energy- focused CPHS use cases, namely a smart building and a smart grid, and evaluated against four established OntoUML-based knowledge engineering methodologies. Results indicate that TKTOnto demonstrates stronger suitability for causal-temporal knowledge elicitation and modelling in the studied energy settings under the selected evaluation criteria. This work contributes a repeatable, methodology-driven approach to structuring causal knowledge in OntoUML for energy-focused CPHSs, supporting explainability and knowledge transfer in environments characterised by dynamic behaviour, human involvement, and expert- dependent knowledge.

Keywords: Model Driven Architecture, low code, enterprise ontology, DEMO, UML, Mendix
Astract: With the ever-growing complexity of modern enterprises, and their supporting IT systems, it becomes increasingly challenging to maintain good business-IT alignment. In recent work, we reported on a model-driven engineering approach to transform, (business) semantically rich, DEMO models to low-code software artifacts for the Mendix low-code platform, with the aim to improve business-IT alignment. The latter approach, however, heavily depends on the specifics of the chosen platform. To reduce IT platform dependence, the Model Driven Architecture approach suggests to discern three levels of models of a system: a business-oriented computation independent model (CIM), an (IT) platform independent model (PIM), and an (IT) platform specific model (PSM). In this paper, we present a more refined approach with the aim to increase the extensibility of the existing DEMO to Mendix transformation to other target IT-platforms, while also “opening up” for other CIMs besides DEMO models. The development of this approach is done in multiple (agile) design cycles, in which pimUML, a novel UML profile to express PIM models, is developed and evaluated for preservation of semantics in each transformation step.

Keywords: Conceptual Modeling, Domain Modeling, Modeling Experience, Pain Points
Astract: Despite the widespread use of domain models, the modeling process remains underexplored, particularly regarding the interactions among agents, products, and activities. Building on prior work that identified 16 recurrent moments of dissatisfaction (“pain points”) experienced by modelers during these interactions, this study offers a deeper analysis to clarify the significance of these pain points and support improvements in modeling practice. Through an online survey of 49 modelers, the study provides empirical evidence on the frequency of these moments, the reasons behind the frustrations they cause, and the strategies modelers use to address them. The descriptive analysis offers valuable insights into these aspects, revealing interesting patterns among modelers. These findings have implications for practice and academia, offering a foundation to enhance the modeling experience and improve the value of domain modeling efforts.

Keywords: Model-Driven Engineering, Artificial Intelligence

Keywords: Internet of Meaning, HESTIA, Domain-Specific Modeling
Astract: At first glance, the Internet of Things brings about an expectation for users (be it individuals or organizations) to interact with the many Internet-connected “things” in a natural way while also enhancing everyday work and life. The emergence of smart cities, and smart homes, also fuels the need for a broad audience to interact with the Internet of Things in a natural way. In current practice, however, users are confronted with the need to negotiate a complex landscape involving a myriad of protocols, standards, and work-arounds to integrate “legacy” devices, etc. We contend that users should not have to think about their world in terms of specific sensors, actuators, gateways, and protocols but rather in terms of room temperatures, the desire to increase the temperature in the living room, the concern that the plants in the garden are watered on time, etc. This creates a need to bridge this gap by creating a semantically meaningful layer of abstraction on top of the sensors and actuators that make up the “device and protocols oriented” Internet of Things, to create an Internet of Meaning. To this end, this chapter reports on the HESTIA framework, which combines: (1) An abstraction of the implementation details pertaining to, e.g., different protocols, standards, etc. (2) A domain-specific (conceptual) modeling framework in terms of which “things” can be captured in a way that is meaningful to the domain at hand (3) Based on this, a domain-specific language that is understandable by the user, enabling users to define control/behavioral rules in terms that are meaningful to them The presented HESTIA framework will be illustrated in terms of examples in the context of home and garden automation. Though such application contexts seem less challenging and complex than industrial Internet of Things applications, the variety of devices and protocols and distance between users and the technical details are often larger than in the case of industrial Internet of Things.

Keywords: Data valuation, data value, MACBETH, MCDA, Multi-criteria evaluation
Astract: Determining the value of data products remains a challenge for enterprises and academia, despite the growing recognition of data as a strategic asset across their business operations. This complexity arises from varying definitions of data value, diverse stakeholder perspectives, and the interdisciplinarity of data valuation. To address these challenges, this study develops a multi-criteria evaluation model based on the MACBETH approach to help Galp Energia, a Portuguese energy company, assess the value of data products in its renewables division. The developed model incorporates seven criteria across an enterprise architecture’s business, data, and application/technology layer, providing a comprehensive assessment of five data products. The study contributes to the literature by proposing a tailorable data valuation approach that may be applicable to other industries. Beyond quantifying the data product value, the resulting model serves as a managerial tool to support data-driven decision-making. The model is constructed using a robust approach and overcomes the limitations of existing models, such as oversimplification and practical implementation challenges. Additionally, it fosters interdisciplinary collaboration between research and industry. Future research directions include using the model as a foundation for developing modular data valuation frameworks, exploring its application across sectors, and integrating cross-sector benchmarks.

Keywords: AI Interpretability, Embedding Space, Internal Representations, Multimodal Data, Semantic Event Matching
Astract: Traditional process mining relies on symbolic event logs that represent activities as discrete labels, often overlooking the rich contextual and semantic nuances found in real-world data such as textual reports, visual records, or sensor outputs. In this paper, we propose a paradigm shift: using the internal representations of AI models—embedding spaces learned from data—as the foundation for process mining. Our framework performs both process discovery and conformance checking directly in these continuous vector spaces, enabling the detection of semantically similar yet lexically divergent events. We evaluate our approach along three dimensions: (i) whether embedding-based discovery maintains or improves accuracy over symbolic baselines, (ii) whether multimodal sources such as video and audio can be processed as unified embeddings for mining purposes, and (iii) whether conformance checking in embedding space enables alignment across noisy or semantically perturbed traces. By treating AI’s internal representations as a novel form of process evidence, we show how process mining can move beyond traditional logs and unlock deeper, semantically enriched interpretations of real-world workflows.

Keywords: Digital Twins, Building Renovation, Added Value of Digital Twins
Astract: Driven by Europe’s ageing building stock and the EU’s sustainability targets, renovation of existing buildings has become increasingly important. Building renovation typically comprises three types of activities (operations & facility management, condition assessment & monitoring, and transformation planning), each involving key decisions that can benefit from advanced decision support systems such as digital twins (DTs). DTs carry the promise of improved decision-making about, as well as the monitoring and understanding of, the twinned entity. This paper investigates how DTs can support, and add value to, the decision-making involved in building renovation and how the added value of these DTs can be assessed and (ideally be) quantified. In doing so, this paper focuses primarily on operations & facility management-related activities. By connecting use cases with quantification methods, this paper suggests pathways to evaluate and prioritise DT investments in the context of building renovation.

Keywords: Business process simulation, Video generation, Process modeling
Astract: Video generation models have opened new opportunities for simulating business processes through realistic visualizations. However, current video generation approaches often fall short of capturing the inherent dynamics and structure of business processes and tend to produce inconsistent simulations that lack the rigor provided by formal process models. To address these limitations, we introduce a novel method termed Petri Net structure-driven video generation, which integrates the inherent structural information from process models to tailor video simulations more closely to the dynamics of business processes. We explore multiple strategies for this tailoring, including i) the use of domain knowledge-rich prompting, ii) a storyboard employing image references extracted from process evidence data, and iii) generated image references informed by process models. We evaluate our method across diverse domains, and demonstrate that the Petri Net structure-driven approach improves the perceived usefulness and consistency of the simulated video, marking a step forward in the use of generative AI for more realistic business process simulation.

Keywords: API, LLM, Microservices
Astract: Linking APIs to the business functions they implement is crucial for handling software operations, especially during recovery from disasters or outages. In this context, the speed and accuracy of operators in linking them impact response time during mission-critical operation activities. Besides, this linkage is essential to designing preventive actions, such as resilience strategies. Automatic API classification using Large Language Models (LLMs) may simplify and speed up APIs-business function linkage. However, previous studies unveiled the barriers practitioners face when deciding on and adopting LLMs in software engineering (SE) tasks due to a lack of guidance for non-experts. This paper aims to lower barriers to using LLMs by systems operators and site reliability engineers (SREs), focusing on the API classification task in the context of operational activities. Based on three cases from the finance industry, we extracted requirements for LLM usage, and assessed 14 recently released LLMs on this task. Our results demonstrate that LLMs accurately classify APIs using business function targets with an F1–Score of 89.5 for the leading LLM without requiring specific LLM expertise and resource-intensive fine-tuning. Besides, our findings on LLMs’ performance and reliability mark a significant advancement in comparing open and closed-source and general and domain-specific LLMs in an SE classification task. Eventually, our experiments yield practical guidance for implementing LLMs in this context. Artifacts used in and generated by the experiments are publicly available at https://bit.ly/llms4apiclassification.