Cybersecurity analysis touches on diverse interdisciplinary aspects, both technical and business-oriented, and diverse stakeholders from different backgrounds must be supported with a knowledge capture environment that operationalizes the “secure by design” principles while also ensuring semantic traceability of the design decisions, to enable AI-based analysis. This paper introduces a domain-specific modeling method supported by a process-centric visual language (a BPMN extension) to enable a knowledge graph treatment for designing cyber threats mitigations in tandem with business process engineering activities and associated data flows. The treatment is intended to be leveraged by LLM (large language model) services, therefore experimentation leans towards AI integration. The domain-specific modeling language (DSML) hybridizes BPMN and threat modeling, further subjected to model transformations into RDF graphs. A demonstrator is implemented on the ADOxx metamodeling platform and its interoperability/adapters for integration with triplestores and LLMs—this makes possible both SPARQL and natural language queries over the prescriptive diagrammatic designs, as analysis approaches. The proposed method can be used to describe contextualized cybersecurity threats during the design phase of a system or during IT system auditing. The integration bridge with knowledge graphs and large language models enhances the system analysis capabilities by streamlining heterogeneous knowledge flows comprising diagrammatic representation, RDF graphs and generative AI, towards an emerging flavor of model-driven engineering. The method was developed iteratively according to the Design Science framework, with its development activities specialized for the DSML-focused Agile Modeling Method Engineering framework.

Symbolic AI has been facing long-term adoption obstacles and a slow uptake caused by the limited availability of frictionless tooling-that can support not only knowledge engineers, but also novice users and educators, with designing and presenting graph exemplars that can be visually communicated, edited and ad hoc processed. There's still a shortage of tools that can democratize knowledge graph (KG) creation, something that is increasingly needed-firstly by educators trying to discuss examples with novices without stumbling on OWL jargon at the earliest step, but also for more recent integration cases such as (i) GraphRAG, where private KGs are called to augment or ensure factuality of large language model services; or (ii) in search engine optimization (SEO), where SEO practitioners lacking knowledge engineering background must embed Schema.org graph data into their Web content. Most visual KG tools are visualizers of KGs created by other means-either in OWL-centric ontology editors posing high expertise barriers, or converted from available serializations, or lifted from legacy data sources. The few actual KG editors are mostly OWL editors neglecting to support the creation of schemaless graph datasets as well as of flexible combinations of graph data and schema fragments. This paper reports on a Design Science effort adopting metamodeling means, traditionally employed for the engineering of domain-specific modeling languages, toward defining a KG development and integration method that facilitates both the visual design of KG exemplars and their operationalization. We aim to balance a diagrammatic look and feel with machine readability of the semantic content being produced-further streamlined in an architecting proposition for integration with LLM services and for the production of Schema.org graph snippets. The DSML was deployed and evaluated as a tool implemented on the ADOxx metamodeling platform, using RDF and LangChain as mediators that streamline the content toward triplestores and LLM services.

Recent developments in the synergy between Large Language Models (LLMs) and Knowledge Graphs have introduced Graph Retrieval-Augmented Generation (Graph RAG) as an extension to traditional RAG. Graph RAG replaces the unstructured textual 'knowledge base' obtained in the retrieval phase of the RAG with Knowledge Graphs to improve entity identification and traceability. In this paper we present a customized Graph RAG framework for handling natural language queries over a knowledge graph with the help of Large Language Models. We conduct a comparative evaluation of several LLMs in responding to different query patterns on a domain-specific knowledge graph, employing four prompting levels defined by the TELeR taxonomy. The outputs are evaluated against ground truth results obtained by executing the corresponding SPARQL queries within a GraphDB triplestore. The content on which the experiments are performed is the Unified Medical Language System (UMLS) semantic network. The results summarize several strengths and weaknesses of the different LLMs relative to the complexity of selected query types, triple patterns, and prompting techniques involved in our Graph RAG setup.

Large Language Models (LLMs) have emerged as powerful tools for interpreting both unstructured and structured inputs, showing potential for enabling natural language interaction across the Business Process Management (BPM) lifecycle. Despite initial experiments with LLMs in BPM tasks, the practical integration of these models into autonomous, context-aware Artificial Intelligence (AI) systems remains largely conceptual and require detailed capability assessments. Visions of AI-augmented Business Process Management Systems (BPMS) propose agents capable of interpreting process semantics, assisting stakeholders through dialogue and grounding their outputs in formalized process knowledge. Yet, a conceptual disconnect persists between the symbolic representations used in semantic BPM and the sub-symbolic mechanisms of LLMs. This paper presents an empirical study that probes this divergence by evaluating the effectiveness of querying RDF-encoded BPMN models with two retrieval-augmented conversational agents powered by OpenAI's gpt-4.1, interoperating by different means with knowledge graphs maintained on Ontotext's GraphDB. Natural language prompts are hereby designed in line with the TELeR taxonomy, addressing a diverse set of BPMN-related queries. The GPT-generated responses are assessed against a subset of metrics from the Retrieval Augmented Generation Assessment (RAGAs) framework. While the work advocates the use of semantic graphs as a mediator in LLM-powered BPMS environments, it also shows limitations in relying strictly on LLM's process analysis.

The Work Systems Framework (WSF) gained recognition as a work-focused way to describe, decompose and drill-down enterprise systems. Several metamodels have emerged over the years to address a requirement for software tooling and machine interpretability as support for Work Systems analysis and design, i.e. alternatives to the traditional snapshot templates used in early adoption cases. The Work System conceptualization was itself gradually refined with taxonomies of its core notions, even formalized in a machine-readable knowledge graph to investigate possibilities of bridging to run-time operational data. As a natural step forward, this paper presents WSF4ADO, an ADOxx deployment of the WSF conceptualization offering a diagrammatic DSML (domainspecific modeling language) that captures the core work system concepts and the intended drill-down navigation experience, with scripted mechanisms to ease the design experience and element typing, and the ability to run semantic queries over semantic networks of inter-related work systems. Methodologically, the research follows the Design Science research paradigm and Agile Modeling Method Engineering as a modeling tool deployment framework.

Leveraging the power of symbolic knowledge representation, this study compares outcomes of natural language interaction with BPMN model serializations, examining semantic graphs derived from RDF export of BPMN provided by Bee-Up, in contrast to the conventional BPMN XML export from the process modeler of SAP Signavio. By prompt engineering, we investigate the proficiency of certain GPT services of OpenAI in navigating the semantic intricacies of RDF and the structural hierarchy of XML, ultimately illuminating implications for knowledge retrieval. The findings delve into the complexities of querying BPMN representations using natural language, revealing the transformative capabilities of RDF, but also the value of BPMN employed as a schema for procedural knowledge graphs—i.e., shifting away from their traditional role as diagrams or automation configurators. As per the experimental results, the RDF export showcases superior richness for natural language queries as the graph-like structure of visual diagrams is closer to semantic networks than to XML tag structures, carrying implications for Business Process Management.

Recent position papers have proposed that the traditional Business Process Management (BPM) lifecycle must be revisited considering generative AI advances, specifically by investigating how Large Language Models (LLMs) can assist various phases of the lifecycle. Inspired by that call to action, this paper reports on a series of experiments on how OpenAI's GPT-4 responds when querying the content of Business Process Model and Notation (BPMN) diagrams, as potential support for the Analysis phase of the BPM lifecycle. We are particularly interested in how BPMN content—typically available in enterprises that adopted the BPM lifecycle—should be exposed to LLM services, therefore we comparatively experiment with diagrams provided as XML serializations or as tool-specific RDF serializations. This is a comparison between a standard serialization characterized by intricate cross-referencing that compensates for the XML rigid hierarchical structure and the “semantic graph” view of RDF that is open-ended in terms of semantic annotation and can be serialized as sequences of statements that resemble natural language. The quality of the answers is assessed using the Retrieval Augmented Generation Assessment framework.

This is an experience report on the course re-design efforts and learned lessons from the authors' host institution, regarding Conceptual Modeling topics. The report is centered on a Systems Analysis & Design course for the Business Informatics bachelor studies program, accredited in the domain of Business Administration. The course establishes a core modeling-based body of knowledge for business analysts and system engineers, which later diverges into related disciplines at master and doctoral levels. The effort started from a legacy UML and requirements engineering course that used to generate what students perceived as a "Diagrammer" skill profile, and the restrictive conclusion that conceptual modeling methods belong heterogeneously to other disciplines, rather than being knowledge capture and analysis enablers offered by a standalone discipline. Based on experience exchanges within the modeling-focused OMILAB community of practice, and inspired by agendas and re-framings advocated in recent scientific literature on Conceptual Modeling education, the authors have pursued the initiative to redefine the curricular SA&D offer as a layered approach. The new approach balances a diversity of tool practice, multi-perspective conceptualization and knowledge work across multiple abstraction layers and standards. To give a constructivist flavor to the learning path, this is synchronized with the documentation requirements for bachelor theses projects. The effects of this re-design show primarily a shift from the "graphical modeling" focus, resulting in a steady stream of publications now realized by students, emerging doctoral projects aiming for Conceptual Modeling contributions (not only tool usage), and benefits pertaining to a better coverage of Bloom's taxonomy regarding Conceptual Modeling learning objectives and competence.

Software development companies are tackling front-end automation primarily from a software testing perspective or a robotic process automation perspective. Industry solutions still lack a knowledge-focused approach to managing automatable front-end workflows. This paper reports on a Design Science Research treatment employing semantic graphs derived from UX (user experience) flow models repurposed for low-code browser automation. The reported work aims to bridge the interoperability gap between UX design, front-end automation and Knowledge Management, using semantic graphs as means of externalization/combination. We developed a visual designer that captures the human-computer interaction flow in procedural UX descriptions with the help of a domain-specific modeling language. The tool further renders visual models executable through a platform-specific automation script, while at the same time maintaining them as RDF graphs accumulating in a knowledge repository. Execution creates a browser instance that replicates user actions within a limited context. The engine relies on Selenium WebDriver, a popular Test Automation framework, to generate automation out of the procedural UX knowledge captured in visual models. As a technical evaluation we demonstrate the feasibility of a low-code automation scenario using the popular browser-based mail client Microsoft Outlook.

Purpose-The purpose of this paper is to provide an effective and comprehensive solution in the form of a project-specific proof-of-concept for handling movie reservations by a hybridization of knowledge graphs (KGs) and large language model services. Design/methodology/approach-The domain-specific scenario of movie reservations offers an interesting testbed for experimenting with master data management and semantic interoperability. While the authors describe a particular solution (for movie reservation systems), the principles and the proposed approach are domain-agnostic. Findings-The proposed solution is more robust in terms of data quality than existing movie reservation systems: by retrieving some of the information from publicly available open data, they have a guarantee that this information is consistently validated and updated since it always refers to the same globally-identified entity made available partly for marketing purposes, partly for encyclopaedic sharing and business interchange. Research limitations/implications-The research data were limited to only the movie domain. DBpedia is used as a core external data source, and it could affect the system's availability. Another possible hindrance is the limited scope of the unstructured data integration - currently, only a small sample from movie blogs was considered. In addition to these technical constraints (i.e. the domain-specific scope, reliance on DBpedia and limited integration of unstructured content), a further area for improvement is the lack of a formal evaluation of the user impact, which is essential for supporting the generalizability of the findings. Practical implications-By integrating advanced features like personalized recommendations and promotions, our reservation system enhances the overall customer experience from the very first point of contact, making the process of discovering, selecting and engaging with the movie experience in a more seamless and intuitive way. Originality/value-In this study, by pursuing a federated knowledge base approach, they were able to show how the notion of servicescape can be supported by means of knowledge science as an alternative to the more traditional approaches popular in marketing, which mainly focus on environmental factors. The resulted servicescape is meant not for general aesthetically pleasing but to meet the personal needs and expectations.

Advocating a convergence between generative AI and BPMN-based process analysis, this study reports on experiments with multi-modal business process representations. By leveraging the capabilities of the Bee-Up modeling tool for RDF serialization and the standard XML export of SAP Signavio, the report probes into the generative AI ability of BPMN interpretation according to these different serializations. In addition, the deployment of multi-modal AI – that directly processes image inputs – transcends traditional constraints of machine readability of BPMN diagrams. For prompt engineering, we employ a combined strategy utilizing semantic processing offered by Ontotext GraphDB integrated with LLM services from OpenAI, which, applied on RDF representations of BPMN, can push the boundaries of natural language interactions with visual process models. The investigation experiments with the interpretation of BPMN process models through such AI-based user interactions, highlighting possibilities of integrating conversational AI with the Business Process Management lifecycle. Assessments of outcomes are based on the RAGAs framework.

The 38th issue of CSIMQ comprises four articles selected by guest issue editors on topics related to novel decision support methods or model-based frameworks for evolving or evaluating information systems. Design-oriented research is the dominant approach in these works, balancing technical design decisions insights with empirical evaluation cases. Three of the selected articles contribute with decision-support methods or frameworks – for ESG (Environment-Social-Governance) accounting, for democratized decision services, and for information security management. The fourth article revisits UML-based model-driven software development from a new perspective.

Data Flow Diagrams (DFD) have proven effective in designing and analyzing the flow of data in enterprise systems. They serve as indispensable tools for enterprises that are undergoing transition to cloud services. DFDs aid in understanding the current processes, identifying interfaces and integration points that require security measures. This paper reports a Design Science project to mitigate the cyber security threats at the design phase of a system and to perform auditing of an existing system through knowledge graphs. The proposal leverages knowledge gathered from various sources in a knowledge graph to identify semantic relationships and patterns, enabling automated inference, analysis and detection of vulnerability patterns. Furthermore, LLM-based (large language models) capabilities transform data management details captured as Data Flow Diagrams (DFD) into knowledge graphs for semantic querying and improved decision support.

The hype of Large Language Models manifests in disruptions, expectations or concerns in scientific communities that have focused for a long time on design-oriented research. The current experiences with Large Language Models and associated products (e.g. ChatGPT) lead to diverse positions regarding the foreseeable evolution of such products from the point of view of scholars who have been working with designed abstractions for most of their careers - typically relying on deterministic design decisions to ensure systems and automation reliability. Such expectations are collected in this paper in relation to a flavor of systems engineering that relies on explicit knowledge structures, introduced here as “semantics-driven systems engineering”.The paper was motivated by the panel discussion that took place at CAiSE 2023 in Zaragoza, Spain, during the workshop on Knowledge Graphs for Semantics-driven Systems Engineering (KG4SDSE). The workshop brought together Conceptual Modeling researchers with an interest in specific applications of Knowledge Graphs and the semantic enrichment benefits they can bring to systems engineering. The panel context and consensus are summarized at the end of the paper, preceded by a proposed research agenda considering the expressed positions.

This issue of CSIMQ includes five articles selected on the topics of complexity management and knowledge management in enterprises from various perspectives and towards diverse goals: the economic and societal perspective seeking strategy fulfillment with the help of enterprise architecture management, the operational performance perspective, which requires grounding management decisions on data insights, and the human aspect perspective to facilitate collaboration, creativity and innovation. Methodologically, the issue reports a mix of conceptual modeling and analytics approaches, suggesting an emerging requirement to balance or alternate, in complexity management, abstraction-based analysis, and data-oriented analytics.

The paper reports on a Knowledge Graph (KG)-based Data Fabric project in the authors’ host institution. The project aimed to lift legacy relational databases that are spread around the institution to an enterprise-grade Ontotext GraphDB server and to provide a visual data navigator to decision-makers. The visualization takes a particular technical approach to be explained in this paper. Complex SPARQL queries are devised not only to retrieve the data to be visualized, but also to gradually transform legacy data towards visual graph structures tailored for diagrammatic navigation through a contextual sliding window. This raises a specific requirement for graph transformations that are valuable as reusable patterns that are more comprehensible compared to the common visual tools we encountered for KG navigation. Therefore the work belongs to the artifact-building genre advocated by Design Science and suggests the importance of a dedicated research direction that should account for the distinction between the semantics of what is stored in the KG and the semantics of visual constructs, of how things are graphically presented – with the help of metadata labelled here as “visualization semantics”, a viewpoint that is well established in the field of diagramming but ignored for a long time by the visual interfaces emerging from the KG community.

Our study reports a comparative analysis of natural language interactions with BPMN models, specifically contrasting semantic graphs generated from diagrams by the RDF export of Bee-Up, against the traditional standard BPMN 2.0 XML export from standard-compliant tools (in our case, SAP Signavio Process Transformation Suite). Utilizing varied prompt engineering techniques, the study evaluates the efficacy of GPT-based services from OpenAI in interpreting the nuanced semantic network structures of BPMN-as-RDF and the standard control flow hierarchical decomposition of BPMN-as-XML. Although image-based multi-modal interpretation of BPMN diagrams is also available in such services, our work is motivated by the fact that most BPM systems deliver structured serializations and not images through their APIs; moreover, any data stored in diagrams cannot be scrutinized by computer vision capabilities, being set as annotations in most tools. By exploring both the challenges and effectiveness of utilizing natural language in interacting with BPMN models, the analysis underscores the ability of RDF to mediate semantic richness and open-ended extension of procedural knowledge compared to the closed-world of the XML interchange schemas. Diagrammatic environments are thus encouraged to pursue this as a potential convergence between different means of knowledge representation.