This is an experience report on the course re-design eforts 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 efort 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 ofered 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 ofer 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 efects 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.
This is an experience report of the on-going efort, over the last decade, to redesign the Conceptual
Modeling (CM) educational ofer for the Business Informatics study program in the authors’ host
institution. It is a sister report to [
Therefore, in this paper we zoom in on a bachelor-level Systems Analysis & Design (SA&D) course, a
curricular slot recommended by the joint ACM/AIS competence models [
general landscape of highly fragmented CM education [
The remainder of the paper is structured as follows: in Section 2 the course redesign treatment will be framed as a Design Science Research (DSR) problem, followed by insight into the problem context in Section 3. Section 4 describes the evolution stages of this curricular re-design from the legacy stage to the current one, further detailing rationale and learning outcomes based on a revised Bloom taxonomy. Section 5 enumerates the teaching tools and key enablers. Section 6 comments on related work and Section 7 provides theoretical connections to educational theories. The paper concludes with a SWOT evaluation summarizing some success indicators and recommendations.
Design Science Research (DSR) being an appropriate frame for proposing prescriptive treatments to
contextualized problems, we remain anchored with this proposal in the position that was formulated
in [
Enhance the CM curricular ofer in the host institution for the field of Business Information Systems (problem context)
. . . by treating it with a course redesign that can position CM as a standalone discipline ofering diverse viewpoints and traceability analysis paths (treatment)
. . . to facilitate the emerging mediation and viewpoint integration roles of CM (treatment service/requirements)
. . . in order to support system analysts who need multi-perspective design traceability and semantic competency going beyond diagramming/visualization – i.e. the legacy focus of the preexisting course design (stakeholder goals)
This was further drilled down into epistemic goals that drive this efort, derived from requirements from upper levels courses (master and PhD) that rely on certain CM foundations:
R1. To reveal a diversity of modeling languages in relation to diversifying purposes, thus showcasing
how CM ofers instruments to various Business Informatics disciplines - some leaning on business, others
on information technology, or alignining the two. We frame this diversity of purposes according to the
Purpose-Specificity framework [
R2. To switch the dominant "Diagrammer" skill profile - inherited from the legacy course - to a
novel knowledge work skill profile, previously advocated under the label "Digital Innovator" in the
educational agenda of the OMILAB network [
R3. To switch the legacy perception - that the main purpose of modeling is to obtain some diagrams
towards the understanding that equally relevant is the modeling cognition, i.e. the underlying efort and
side-efects - situational comprehension, sense-making, gap filling - that a system analyst goes through
while creating diagrams. This efort includes for example identifying decisions and corner cases in
case management, identifying dependencies towards architectural completion, comparing alternate
chronologies in a process design, ultimately ensuring the right level of detail becomes available for
subsequent retrieval and traceability, in external systems relying on the model contents. This also
gives occasion to discuss the ChatGPT-driven irrelevance of output-based student evaluation (even
encouraging assessment automation in past literature [
R4. The above mentioned goal of "retrieval and traceability" (in external model-based systems)
implies that model querying demonstrations are provided as early as possible to students - e.g. starting
with process querying methods [
To satisfy all the above requirements, an overarching line of argumentation must encompass the
course progression and learning objectives, summarized in several stages depicted in Figure 1. This
progression starts from highlighting the diversity of CM methods and standards, to being able to choose
between them in relation to purpose and desired competence. On the right side of the figure, there
is a suggestion of how this flow advances to the master program. In the master program, we can
directly step into DSML (domain-specific modeling language) engineering as a means of expanding
the competence/scope of any modeling standard, and therefore the range of competency questions
that diagrams can answer. This is based on educational tooling that easily demonstrate the focus on
model contents semantics and metadata, and not only visualization. Exemplary topics that take this
idea further and demonstrate diverse applicability are studied in the English-language master program
of Business Modeling and Distributed Computing4:
• Process queries, simulation and mining (in BPM context)
• Enterprise Architecture reporting (in EAM context)
• Low-code (in Model-driven engineering and Internet-of-Things context)
• Digital Twins (in Model-driven Knowledge Graphs context, primarily Knowledge Graphs derived
from BPMN models as advocated in recent experimental demonstrators [
These topics have been picked not only because of a desire to balance business and informatics topics, but also due to their reliance on relational traceability and model navigability, thus showcasing a diverse applicability layer on top of the CM foundations established by the course hereby discussed. We found it critical to introduce model queries from the earliest meetings (briefly showing demonstrators, e.g. 4https://econ.ubbcluj.ro/programe/bmdc/ report generation) as this is perhaps the one unexpected feature that easily catches attention, sparks curiosity and stimulates lateral thinking for those students stepping into the course with the legacy mindset of the Diagrammer profile. The legacy profile is sometimes inherited from past generations, or from Web presentations that superficially introduce CM as a "drawing" activity, thus maintaining for novices a disconnect from subsequent model-based activities and implications. Consequently, our choice of educational tooling (see Section 5) aims to avoid simple diagramming canvases where visual sketching and inspection are the only tasks supported.
Some points of discourse deviation are also shown as downwards arrows in Figure 1, where the
potential involvement of Generative AI in CM activities is suggested. This has become a natural
curiosity for students, and the CM-GenAI interplay is increasingly prevalent in literature – from
investigating how GenAI can serve the Business Process Management lifecycle [
CM education has always faced dilemmas in the search for an ideal balance between practitioneroriented relevance and formal grounding, between tooling and standardized principles. Challenges can also be case-specific, such as in the regional context where this report originates – an IT outsourcing hub that advertises cost savings, traditionally considering CM eforts to be a costly overhead of little pragmatic relevance, for a number of diverse reasons: • Design decisions are often taken for granted or inherited from legacy systems being maintained by the outsourced service provider. Long-term loyal partnerships lead to intimate understanding of the client-side business processes and legacy architecture – already available in past documentation, or shared through knowledge management approaches (wikis, Confluence etc.); • IT services tend to be specialized and sometimes tackle only small parts of the software development lifecycle – software testing, front-end "face lifting", DevOps etc. Workers involved in such isolated concerns are often not interested, or do not have access to a system-wide view of the enterprise system they contribute to. The argument of "agility" is occasionally brought forth to justify this, even when it leads to lack of awareness of certain dependencies; • Business/process analysts have been absent from software projects for many years, until some customers or certification requirements imposed their presence on the outsourced side. A typical example is the sudden success of RPA (Robotic Process Automation), where outsourcing providers quickly set up tool-specific (UiPath, Power Automate) developer teams deploying RPA implementations in the absence of a Business Process Management culture. This has lead to widespread improvisation of process description methods neglecting BPMN and ad-hoc process analysis methods based on spreadsheets and visual inspection.
On the other hand, recent crises waves in the IT industry – from the pandemic to several stages of automation – revealed new distinctions in capabilities where CM can play a role, even if not explicitly labelled as CM (practical work calls it "process mapping", "low-code development", "visual maps" or, simply, "diagramming" or "design"): • Distinctions between reactive/reflexive companies, preoccupied strictly with identifying opportunities to ofer highly specialized services and human resources, as opposed to companies developing their own products, even meta-products whose design decisions must be tailored to diverse business processes and domain specificities; • Distinctions between companies that lost their outsourcing partnerships (because of e.g. backshoring in the context of RPA and AI), as opposed to companies that managed to turn themselves into automation providers, including associated consultancy "centers of excellence"; • Some companies indirectly acquired CM talent and capabilities after facing impositions from customers or projects to demonstrate availability of knowledge work roles (process analysts, knowledge managers, enterprise architects) or certifications (ITIL, TOGAF etc.); • Some companies shifted from using commercial low-code platforms to developing their own low-code environments to improve productivity and expand development capabilities. This took them inevitably on a model-driven engineering path.
On the academic side, the curricular ofer accommodated such shifts over time in a mostly ad-hoc
manner, reacting to hypes, adopting tool-centric tutorials and not really pursuing an explicit CM
competence model/skill profile. Some disciplines always incorporated CM tooling (ER in Databases,
class diagrams in Object-oriented Programming), while others followed industry trends (e.g. RPA as a
tool-centric tutorial in the absence of any Business Process Management or Enterprise Architecture
contextualization). CM methods and tools have spread between disciplines in a highly fragmented
manner, which is in line with the conclusion of the literature survey in [
As a group of educators responsible with managing a stream of courses – starting from SA&D at bachelor level (the focus of this paper) and later diverging at master level into BPM, EAM, RPA, MDE and low-code IoT – we engaged in a longitudinal curricular redesign and streamlining that positions CM as a standalone discipline with a rich ofer of tooling and methodological enablers to a diversity of disciplines in Business, Informatics, or their convergence.
After first introducing, for a research-oriented master program (around 2015-2017) several teaching
artifacts involving DSMLs and experimental interplay between knowledge graphs and diagrammatic
modeling (see Section 6 and the sister report in [
The Legacy Stage: Stable curricular content has been maintained for close to a decade before
we started getting involved in this course. It used to be based on a selection of core UML diagrams
(use case, activity, state, class), complemented by data modeling (ER), and book-ended by informal
visualizations (e.g. the Fishbone diagram) and software project management considerations on Waterfall
and various Agile flavors. This was reasonably aligned with the joint ACM/AIS recommendations
for SA&D competences [
Stage 2. Diversification of Conceptualization: The new approach allows to navigate several standards and multiple tools, enabling discussion on model competence overlapping and complementarities. Subtly, this set the grounds for later introducing model queries to highlight the value of formal modeling for machine-readable enterprise knowledge structuring. Parts of UML and data modeling, already covered by prior disciplines, are only briefly revisited; instead, the focus is shifted to business process modeling and architecting, designing systems primarily around chronologies (workflows, data lfows) and architectural dependencies, which tend to become prominent requirements in the local industry, especially since the uptake of Robotic Process Automation.
We thus introduced the BPMN ecosystem, including coverage of complements such as CMMN, DMN, DFD. Further down the road, workflows are contextualized by enterprise architecting, where traditional UML diagrams (use case, deployment, components) are compared to the more integrated approach of Archimate. This helps stress the multi-viewpoint modeling needed to manage architectural complexity and enterprise layering. Archimate is not presented in full due to time constraints, so we focus on (a) comparing its core layers to BPMN and UML, and on (b) wrapping around those core layers the Physical deployment of a business, and the Motivation layer (later, at master level, Archimate is picked up in the context of TOGAF5 and the remaining layers are added). In close relation to the Motivation layer of Archimate, informal diagrams are introduced as possible input for motivation modeling. These informal visual templates include Fishbone, SWOT analysis and Business Model Canvas6 to capture high-level business model summaries as a starting point for architecting an enterprise or its information systems.
Stage 3. Moving Beyond Diagramming: One major acceptance obstacle in the UML-focused legacy course, also partly manifesting in the updated version described as Stage 2, was a dominant perception that the course is about various kinds of "drawings" contributing to a Diagrammer skill profile. In students’ feedback, this profile was associated with several demotivating assumptions that needed to be addressed: • that the Diagrammer work is mostly motivated by a pedantry in complying with graphical standards, when informal diagramming (Miro, Draw.io etc.) is just as good for intuitive communication – local industry practice being actually dominated by informal approaches rather than by formal standards, under the pretext of Agile development moving away from the "UML days"; • that the Diagrammer goal is to obtain some diagrams, and this efort can now be delegated to Generative AI; moreover, large language models allow for the Diagrammer to be replaced entirely by natural language "storytelling" as means of both knowledge acquisition and representation, i.e. reducing diagrams to a historic compromise.
Several countermeasures were necessary to push through such assumptions and defuse the legacy
Diagrammer competence profile. This is based on stimulating a constructivist CM education approach
5https://www.opengroup.org/togaf
6https://www.strategyzer.com/library/the-business-model-canvas
by showcasing from the earliest courses demonstrators or cases where those assumptions do not hold,
where multi-perspective enterprise modeling and systems engineering gain value:
• The legacy Diagrammer competence profile, based on small isolated diagramming exercises
that mostly stress syntactic correctness, used to render opaque all CM cognitive eforts –
i.e. knowledge acquisition, sense-making, drill down and gap-filling. These are the basis of
understanding a system-under-study and documenting its As-Is vs. To-Be vision gaps. Diagrams
are presented as a proof that the cognitive efort took place, that the situational understanding
was achieved – thus delegating the efort of "drawing", is not the same as delegating situational
understanding, identification of types of things and dependencies manifesting between them.
Students are tasked not only with isolated diagramming exercises, but with a complex software
project where their software development eforts run in parallel with enterprise modeling, business
analysis and business-IT alignment based on CM methods. This is enforced through Bachelor
thesis guidelines, as the final thesis project is streamlined with the SA&D course to allow for
complexity management concerns to arise, finally justifying the aforementioned cognitive eforts
- as a learning experience clearly distinct from graphical content creation.
• As a communication device, students are shown as early as possible the empirical efectiveness
of visual inspection of a complex diagram, compared to the same situation expressed via
storytelling. It is advisable to organize exercises where question answering accuracy from students
vary between having to inspect a text and a diagram of the same situations, especially when the
questions involve multi-hop navigation of connectors/dependencies, more dificult to grasp from
pages of linear text;
• Afterwards, this visual inspection is replaced by model queries to highlight how the apparent
"visual pedantry" is actually an efort of reducing ambiguity for both humans and machines, i.e.
allowing queries to deterministically distinguish and retrieve desired element types in support of
automated reporting or model-driven processing. Most demonstrated model queries are based on
the standard XML serializations available as export formats in all modeling tools involved – i.e.
XPath7 is shown as the most general retrieval approach applicable across all studied standards
and tools, while also mentioning some tool-specific ones (e.g. the ADOxx query language 8
applicable to BPMN, UML, DMN, UML in the Bee-Up tool9) and others reported in the scientific
literature [
At this stage, diferent model competencies, each illustrated by key concepts - chronology (of actions, events), data (structuring and requirements), architectural dependencies – are navigated by queries tailored for all modeling standards, which helps establish quickly that diagramming is only a superficial visual layer wrapping around the mediator role of CM, with its knowledge structuring and traceability concerns. Some examples of queries that can be easily associated with pragmatic use cases are: – a list of BPMN user tasks to be turned into Jira tasks – a list of DMN data inputs to be turned into data requirements for a decisional API – a list of Archimate elements that influence a specific Motivation layer object. – generalized template applicable across standards: a list of elements of a specific type, involved in a specific relationship or relationship chain.
Recurring query patterns are discussed - first as natural language questions that students must be
able to formulate in order to demonstrate their understanding of the scope/competence of each type
7https://www.w3.org/TR/xpath-31/
8https://www.adoxx.org/documentation/75_adoxx_development_languages/01_AQL.html
9https://bee-up.omilab.org/activities/bee-up/
of model. An early observed obstacle is the traditional understanding of queries (as "data queries, e.g.
SQL) that comes with a tendency to formulate questions about potential model traces or execution-time
instances (who are the employees that executed task X?), instead of querying visible model contents
(to what participant lane/pool does task X belong?). After this distinction is clarified, the question
patterns can be gradually translated into executable queries while diversifying the technical means to
execute them - we insist on XPath queries over XML serializations at bachelor level, after showcasing
a few tool-specific features. However, as shown in Figure 2, later at master level this advances to
more specialized query methods (RDF-based queries as allowed in the OMILAB ecosystem toolkits
[
Stage 4. Enforcing CM Constructivism: Bachelor Thesis guidelines were developed to enforce the reduction of narrative text in theses (lately susceptible to GenAI contamination) and to stimulate a switch towards presenting and arguing for design decisions and business rationale with the help of CM tools and standards. This varies with the type of thesis (also allowing combinations):
Type 1. Software development projects (dominant, about 80% of bachelor theses) must be framed in the written thesis by a business model and enterprise architecture where the developed software could play a role. Such documentation starts from high-level informal considerations - Fishbone causality and unstructured mind maps ideation -, refined into Archimate Motivation and Business Model Canvas, then turned into business processes and finally going through granularity levels of system architecture. Students have the choice to pick their favourite architecting language (UML, Archimate, DFD) and business process perspective (i.e. between strongly structured workflows and case management, possibly involving DMN if business rules play a role). Contrasting between the As-Is vs. To-Be models is important to highlight the business transformation that the proposed software artifact brings.
Type 2. Case studies from the students’ workplaces, typically pertaining to the adoption of new work procedures or enterprise systems is second category of theses; here, CM serves mostly for As-Is vs. To-Be comparisons and some characterization of existing software. This type of thesis is a minority, although not rare (15-20%).
Type 3. Scientific research makes less use of CM methods - mostly to reflect empirical data processing concerns (ata processing pipelines and data structures), unless CM itself is at the core of this research. This type of thesis is, however, a rarity at bachelor level, as it requires scientific methods only introduced in master studies.
Table 1 summarizes learning objectives that are the basis for examination in relation to the revised
Bloom taxonomy of learning levels proposed in [
The ability to recall and give informal defini- The ability to recall and give informal
defitions on concepts present in the investigated nitions on the building blocks of a modeling
standards – e.g. BPMN gateway, CMMN method, according to established DSML
encase, DFD entity (e.g. vs ER entity), Archi- gineering methodologies e.g. Agile
Modelmate business function etc. ing Method Engineering phases [
The ability to describe in natural language The ability to describe in natural language a situation depicted in diagrams with non- a metamodel depicted as a class diagram, explicit labeling, i.e. enforcing the recogni- and to instantiate a case as a mock diagram tion of model elements, rather than extrap- conforming that metamodel. olation from textual labels.
The ability to "translate" a piece of natural The ability to draw the domain-specific text into a diagram of a specified type, pre- metamodel capturing all relationships and serving as much as possible of the details types involved in a natural language descripprovided in the textual description. tion of a situation.
The ability to assess whether a competence The ability to assess whether a competence question can be satisfied by the contents of question can find its answer in models crea diagrammatic model, with respect to the ated accord-ing to a given domain-specific scope of the modeling standard being used. metamodel (given a legend a mock symbols). The ability to execute model queries in order to produce a simple report listing model elements that satisfy certain restrictions.
The ability to find mistakes in models, and The ability to assess whether a domainto distinguish between syntactic mistakes, specific diagram deviates from a metamissing information or mistakes relative to model it is supposed to be compliant with. the situation being represented.
The ability to "translate" a large-scale situa- The ability to implement a novel modeling tion (the Bachelor thesis business case) into tool that deploys a metamodel designed by a collection of adequate multi-perspective students. The ability to use it to describe diagrams of diferent types, depicting dif- instance situations. The ability to build a ferent aspects of a situation/system - using front-end that demonstrates the retrieval of the preferred tooling, while ensuring that model elements at run-time. software implementation is in synch with the diagrammatic design decisions.
The tooling involved in the new SA&D course design is an assortment of diverse educational tools available to students: • On one hand, we employ SAP Signavio12 (for BPMN, CMMN, DMN and Archimate), Archi13 (for
Archimate) and Visual Paradigm14 (mainly for DFD but it also serves many other languages);
• In parallel, we employ toolkits from OMILAB’s Digital Innovation environment [
Between the dominant tasks - modeling, model analysis/querying - other tools may be shown as teasers for the master program: BPMN automation in Tasky Cloud17, Archimate analysis reports in ADOIT18, ADOxx19 to expand modeling standards and querying possibilities towards DSMLs, UML class code generation in Bee-Up, DMN and Process simulation etc. However these are brief intermittent detours, as practical tasks remain focused on gaining familiarity with the diversity of modeling grammars and standards, understanding their specialized competence and overlaps, gaining hands-on experience with diverse tooling and being able to extract model information via queries. The rationale for the selection of tools, besides their open availability to students, is: • to avoid creating the impression of a tool-centric tutorial (e.g. "this is the StarUML course", a perception detected in the legacy stage of the course); • to enforce a clear distinction between a modeling language/standard and tooling, raising awareness of diferences and overlaps; • to allow any student to pick preferred tooling for their Bachelor theses; • to avoid diagramming-only tools that are not able to demonstrate any model-based feature outside visual formatting (a critical point in defusing the Diagrammer focus).
As background priors, we refer to some of our earlier reports that provided insights to teaching cases
and courseware that was experimentally employed at master and doctoral level before taking over the
bachelor-level content: the "cooking recipe modeling method" gives insight to the constituents of a
minimalist rudimentary workflow language [
Those enablers and teaching strategies, however, required a foundation established in the bachelor level CM-focused course presented in this paper. Therefore, the course takes a generalized approach to model queries relying on the generalized availability of XML serializations, leveraging existing XPath skills gleaned from tasks in other courses. The ability to use queries to navigate the dependencies and decompositions captured in models becomes thus a unification concern across all CM standards or tools, as a more powerful and deterministic model process approach compared to solely visual navigation. It is also a key message against perceiving CM as a fragmented visually-focused practice, reframing it as a knowledge refinement and retrieval practice.
For CM education research this proposal opened new possibilities for empirical studies - e.g. with
Generative AI model queries across diferent model types, patterns and formats [
Recent scientific agendas have also proposed a reframing of CM education as a standalone discipline
ofering technological and methodological enablers to other disciplines, as opposed to the traditional
fragmentation. The extensive bibliometric analysis of [
Constructivist learning [38] recommends that students should derive knowledge by their own
construction efort rather than by direct assimilation of content or imitation tasks. The CM education
research survey of [
In our case, students end up performing a guided construction efort mainly for their bachelor thesis, where the diversity of modeling languages and tools are used in a coordinated manner to describe bachelor projects from multiple perspectives, both business and software-oriented. Most projects are software systems that must be not only developed but also contextualized by a (real or fictive) business model, business processes and enterprise architecture. The SA&D course runs in parallel with this construction efort by gradually introducing the diferent model types and competencies involved there. The course builds on several prior modeling skills presented in earlier disciplines in a fragmented manner (mainly data modeling presented in Databases courses as part of the relational DB design process) and on XPath as generic model navigation and retrieval means - to stress that models are not an end in themselves, but must interoperate in larger systems as semantically-prescribed mediators.
Through its ramifications to other already mentioned master program disciplines, the course also
follows the spiraling strategy recommended by J. Bruner [40] - i.e. a recurring revisitation of the same
topics throughout the curriculum. XPath queries later diversify into more specialized model query
methods [
Strengths. The main goal of this longitudinal course redesign efort is to enable, in the spirit of the humboldtian education model, a capability of producing CM scientific output authored by students and at the same time a constructivist approach to CM education. The students become able to create a model-based knowledge repository for their system engineering outcomes. This also makes them more prepared for junior research work in projects and for following a CM-centered PhD program – which only became possible recently; conference publications based on student theses or their research project work are now regular - we are tracking them at https://econ.ubbcluj.ro/omilab/publications.php, with typical venues being BIR, CAISE workshops, KES, ECIS, ISD, often in tool/forum sections due to the practical focus of student work. Institutional projects became possible, employing student talent now familiar with purpose-divergent CM methods, including DSMLs; one recent example of a successful institutional project involves a student-developed DSML for managing process design decisions in Robotic Process Automation projects [41].
Weaknesses. One major shortcoming for an SA&D course is the absence of SysML, which we postponed in order to wait for the new, radically re-designed version announced for 2025. By reducing current UML coverage, already partly represented in a parallel software engineering courses, we can bring in the new vision SysML that also shares our concern for models as logical knowledge structures not necessarily visualized diagrammatically.
Opportunities. We have yet to look at the emerging opportunities brought by new versions of SysML
and Archimate, in relation to the noticeable interest in the convergence of modeling languages and
RDF/OWL, see the knowledge graph taskforce recently established by OMG21. This has the potential to
standardize eforts regarding ontological commitment [42], reasoning and semantic querying [
We are also interested in involving the less formal practice of Design Thinking, as its bridging to CM has been at the core of the OMILAB Digital Innovation environment, specifically in the Scene2Model toolkit [43]. Design Thinking is a major preoccupation in our local industry, typically framed as a marketing practice rather than a CM preliminary efort of knowledge acquisition and ideation.
Threats. The new course design moves away from the joint ACM/AIS curricular recommendation in
[
It is not clear how the industry and labor market assimilate the new skill profile emerging from this course re-design. It is too early for longitudinal assessments, but early signals show increasing interest on the companies’ side in conceptualization and abstraction skills - especially as entry-level coding jobs start being delegated to GenAI products. Human-driven conceptualization is even in our outsourcing-dominated industry heavily involved in Product Development (from Design Thinking to Process Mapping), Robotic Process Automation (in the context of As-Is vs. To-Be business process analysis), Low-code Development (in the context of low-code platforms use and customization) and we hope for future longitudinal surveys to confirm the value of our multi-perspective CM ofer.
As recommendations for potential adopters of similar strategies we formulate the following guidelines: introduce diverse modeling standards that show both semantic overlaps and distinctions; clarify their competency/scope boundaries regardless of visual customizations allowed by some tools; show the oficial specifications to reveal the importance of machine-readable serializations and the XML schemas governing them; bring forth a notion of model queries reflecting that of competency questions; pick educational tooling that easily demonstrate them through querying features or some model-based reporting (XPath may serve as a fallback demonstrator); employ standards that have reasonable visibility - start from UML, BPMN, Archimate, and expand to more specialized languages or DSMLs as time allows, maintaining focus on how this reflects in expanding competency and querying possibilities. 21https://www.omg.org/ekg/ 22Bee-Up has been a long term promoter of the RDF export for a numbering of modeling languages and model types, thus allowing the model query to be demonstrated via SPARQL, see https://bee-up.omilab.org/activities/bee-up/
The author(s) have not employed any Generative AI tools. of the 14th Internationale Tagung Wirtschaftsinformatik 2019, 2019, pp. 1998–2002. URL: https: //aisel.aisnet.org/wi2019/specialtrack03/papers/7. [35] D. Bogdanova, M. Snoeck, Domain modelling in bloom: Deciphering how we teach it, in: Proceedings of PoEM 2017, volume 305 of LNBIP, Springer, 2017, pp. 3–17. doi:10.1007/ 978-3-319-70241-4_1. [36] K. Rosenthal, B. Ternes, S. Strecker, Learning conceptual modeling: structuring overview, research themes and paths for future research, in: Proceedings of ECIS 2019, Association for Information Systems, 2019. URL: https://aisel.aisnet.org/ecis2019_rp/137, paper 137. [37] I. Maslov, S. Poelmans, Advancing the BPMN 2.0 standard with an extended animated notation: A research program for token-based process modeling education, in: Proceedings of BIR 2023 Workshops and Doctoral Consortium, volume 3514, 2023. URL: https://ceur-ws.org/Vol-3514/ paper92.pdf. [38] V. Richardson, Constructivist teaching and teacher education: Theory and practice, in: Constructivist Teacher Education: Building a World of New Understandings, Routledge, 1997, pp. 3–14. [39] U. Frank, Conceptual modelling as the core of the information systems discipline–perspectives and epistemological challenges, in: Proceedings of AMCIS 1999, 1999, pp. 695–697. URL: https: //aisel.aisnet.org/amcis1999/240. [40] J. S. Bruner, The Process of Education, Harvard University Press, 1960. [41] Ş. Uifălean, R. A. Buchmann, Low-code browser front-end automation using RDF graphs and a domain-specific language for UX representation, in: Research Challenges in Information Science (RCIS) 2025, volume 547 of LNBIP, Springer, 2025, pp. 140–155. doi:10.1007/ 978-3-031-92474-3_9. [42] G. Guizzardi, G. Wagner, R. Guizzardi, Towards ontological foundation for conceptual modeling: The unified foundational ontology (UFO) story, Applied Ontology 10 (2015) 259–271. doi: 10. 3233/AO-150157. [43] C. Muck, S. Palkovits-Rauter, Conceptualizing design thinking artefacts: The scene2model storyboard approach, in: Domain-Specific Conceptual Modeling: Concepts, Methods and ADOxx Tools, Springer, 2022, pp. 567–587. doi:10.1007/978-3-030-93547-4_25.