Skip to content

Related Work: ArchiMEO — Enterprise Ontology (FHNW)

ArchiMEO is an enterprise ontology developed at FHNW (University of Applied Sciences Northwestern Switzerland) starting around 2010, primarily by Andreas Martin, Barbara Thönssen, Knut Hinkelmann, and collaborators including Roman Brun, Simon Nikles, Sandro Emmenegger, Emanuele Laurenzi, Devid Montecchiari, and Maja Spahic. It is one of the most academically rigorous efforts to formalize enterprise architecture concepts as OWL ontologies, validated through five applied research projects and documented in peer-reviewed publications spanning over a decade.

The name "ArchiMEO" indicates its foundation in ArchiMate ("Archi") plus its adaptation and enhancements to serve as a Meta Enterprise Ontology ("MEO").

Key Publications

  • Hinkelmann et al. (2020) — ArchiMEO: A Standardized Enterprise Ontology based on the ArchiMate Conceptual Model, MODELSWARD 2020, pp. 417–424. DOI: 10.5220/0009000204170424. The primary reference paper describing the ontology structure, development methodology, and five application domain extensions.
  • Hinkelmann et al. (2016a) — A new paradigm for the continuous alignment of business and IT: Combining enterprise architecture modelling and enterprise ontology, Computers in Industry, 79:77–86.
  • Hinkelmann et al. (2018) — Ontology-based metamodeling, in Business Information Systems and Technology 4.0, pp. 177–194.
  • Martin (2016) — PhD thesis on ICEBERG (Interlinked Case-Based Reasoning), which uses ArchiMEO as its enterprise ontology foundation.
  • Hinkelmann et al. (2026) — Ontology-Based Validation of Enterprise Architecture Principles, Applied Sciences, 16(7):3352. The most recent paper extending ArchiMEO with SHACL/SPARQL-based automated compliance checking of EA principles.

The Project

ArchiMEO's stated goal is to serve as a "standardized Enterprise Ontology" providing a shared conceptualization for integrating heterogeneous enterprise data sources. It was validated through five applied research projects: contract management, supply-risk management, experience management (case-based reasoning), workplace learning, and Business Process as a Service (BPaaS).

The project followed the widely-used ontology development methodology described by Noy et al. (2001):

  1. Competency questions as benchmarks — natural language questions the ontology must answer
  2. Practitioner interviews to refine extracted concepts
  3. Existing ontology analysis — TOVE, The Enterprise Ontology, CEO, REA were analyzed but found unsuitable as standards
  4. ArchiMate selection — chosen as the most suitable EA framework backbone given its broad acceptance in practice
  5. Standard ontology engineering steps — enumerate terms, define classes and hierarchy, define properties and facets, create instances

This methodology is more formally grounded in ontology engineering literature than most community-driven efforts, with explicit competency questions and practitioner validation. The academic peer review and multi-project validation provide a level of rigor that is rare in this space.

Ontology Architecture

ArchiMEO is structured as a four-layer architecture:

  1. Top-Level Ontology (TOP) — Generic world concepts (location, time, events)
  2. Enterprise Upper Ontology — ArchiMate as OWL
  3. Meta Enterprise Ontology (ArchiMEO/EO) — Domain specializations and enterprise vocabulary
  4. Application-Specific Ontology — Project-specific extensions

Top-Level Ontology

ArchiMEO includes a custom Top-Level Ontology with concrete domain concepts: Location hierarchy (PhysicalLocation, LogicalLocation), Time hierarchy (TimePoint, TimeInterval, CalendarElement), and Event hierarchy (ForceMajeureEvent, ActOfGod, MenMadeEvent). As stated in the MODELSWARD 2020 paper: "ArchiMate does not provide general concepts, like location or time. Therefore, we introduced a Top-Level Ontology following Bertolazzi et al. (2001)."

This is useful for applications that need to reason about geographic and temporal context — for example, supply-risk management where supplier locations matter. Linked.archi takes a different path here, deferring to standard W3C vocabularies (SKOS, PROV-O, Dublin Core, Schema.org) rather than defining its own upper ontology, which means users bring their own time/location/event concepts when needed.

Enterprise Objects (EO)

The Enterprise Object ontology (~2,769 lines) is the largest module and serves as the integration hub. EnterpriseObject is the root class for all enterprise concepts, and all ArchiMate core concepts are subclasses of it. The module includes extensive business domain content: organizational roles (Employee, ManagerRole, ChiefOfficerRole), business partners (Client, Customer, Supplier), documents (ContractDocument, Specification), business events, products, IT systems, SWOT concepts, and business model concepts.

The EO module is notable for mixing three fundamentally different modeling patterns:

Pattern 1 — Parent classes of ArchiMate elements (upward extension). EnterpriseObject is the root that all ArchiMate concepts inherit from, enabling cross-notation queries regardless of source notation.

Pattern 2 — Child classes of ArchiMate elements (downward specialization). Domain-specific specializations like Person rdfs:subClassOf archi:BusinessActor, Client rdfs:subClassOf BusinessPartner rdfs:subClassOf archi:BusinessRole, bridging BPMN to ArchiMate via bpmn:Activity rdfs:subClassOf eo:BusinessActivity.

Pattern 3 — Individuals (instances). Actual reference data mixed with class definitions: geographic instances (Switzerland, Germany, Japan), natural disaster instances (Earthquake, Tsunami), role instances.

This mixture exists because ArchiMEO was designed for applied research projects where the ontology needed to be immediately usable with real enterprise data. The reference data (countries, disasters, roles) is shared across multiple application-specific ontologies.

ArchiMate Module

Based on ArchiMate 2.1 (2013), the module covers ActiveStructureElement, BehaviourElement, PassiveStructureElement, Service, and Interface hierarchies. Relationships are modeled as both OWL classes and ObjectProperties. All elements are subclasses of eo:EnterpriseObject, creating tight integration with the enterprise vocabulary.

ISO 42010 Module

A dedicated module with 12 classes and 17 object properties covering Architecture, ArchitectureDescription, ArchitectureModel, ArchitectureView, ArchitectureViewpoint, Concern, Correspondence, CorrespondenceRule, ModelKind, Stakeholder, and System-of-Interest. The Correspondence and CorrespondenceRule classes are particularly valuable — they formalize cross-view consistency constraints that are important for architecture governance. Linked.archi's core ontology integrates ISO 42010 concepts directly rather than as a separate module, covering most concepts but currently lacking Correspondence and CorrespondenceRule.

Business Motivation Model (BMM)

A dedicated module implementing the OMG Business Motivation Model: Vision, Mission, Goal, Objective, Strategy, Tactic, BusinessPolicy, BusinessRule, Assessment, PotentialImpact (PotentialReward, Risk), and Asset. This provides a formal bridge between business motivation and architecture that goes beyond what ArchiMate's Motivation layer offers alone.

BPMN Module

A focused module covering BPMN 2.0 core constructs (~45 classes): tasks, events, gateways, flows, participants, and lanes. Cross-ontology integration is achieved through direct subclass axioms (bpmn:Gateway rdfs:subClassOf archi:Junction, bpmn:SequenceFlow rdfs:subClassOf archi:Triggering). Linked.archi's BPMN suite takes a different approach — a full XMI-derived mapping (144 classes) with spec-pure ontology and separate alignment files for cross-notation bridging.

Enterprise Model Ontology (EMO)

An integration layer providing explicit, queryable mappings between model types and their constituent elements. This enables questions like "What elements can appear in a BPMN model?" or "Which model types cover the application architecture aspect?"

The Integration Strategy

ArchiMEO integrates modeling languages by making their concepts subclasses of a shared EnterpriseObject hierarchy. For example, archi:BehaviourElement rdfs:subClassOf eo:EnterpriseObject and bpmn:Activity rdfs:subClassOf eo:BusinessActivity rdfs:subClassOf archi:BusinessBehaviourElement. This creates a single unified class hierarchy where cross-notation queries are natural.

This approach has clear benefits:

  • Immediate cross-notation reasoning — a single SPARQL query can find all behavioral elements regardless of whether they came from ArchiMate, BPMN, or BMM
  • Unified enterprise vocabulary — organizations can work with domain-specific terms (Customer, Contract, DatabaseSystem) rather than abstract metamodel types
  • Proven in practice — validated across five research projects with real enterprise data

The trade-off is tight coupling between modules (circular imports make selective adoption difficult) and the mixing of schema-level definitions with instance data. Linked.archi's integration approach uses a shared core ontology (arch:Element, arch:QualifiedRelationship) with separate alignment modules — more loosely coupled but requiring explicit alignment work for cross-notation queries.

Applied Research Domains

The five validated application domains demonstrate concrete patterns for enterprise ontology use:

  1. Contract management — Automatic metadata extraction from documents, linking to enterprise events (bankruptcy triggers contract review). Pattern: external event monitoring → ontology reasoning → automated alerts.

  2. Supply-risk management — Semantic Risk Model with RiskEvent, RiskIndicator, CrisisPhase, WarningSignal. Pattern: bottom-up risk aggregation from atomic indicators using inference rules.

  3. Experience management (ICEBERG) — Ontology-based case-based reasoning using ISO 42010 viewpoints for stakeholder-specific case retrieval.

  4. Workplace learning — Competency profiles, learning preferences, and recommendation system using ontology-based competency matching.

  5. BPaaS — Aligning business requirements (BPMN) with cloud offerings using semantic matching.

These patterns are transferable regardless of the specific ontology used. The supply-risk and contract management patterns are particularly relevant for governance and impact analysis use cases in any architecture knowledge graph.

Principle Validation (2026)

The most recent ArchiMEO paper (Hinkelmann et al., 2026) describes a systematic approach to validating EA principles against EA models using SHACL and SPARQL. This represents ArchiMEO's evolution toward modern semantic web standards — adopting SHACL for closed-world constraint validation rather than relying solely on OWL reasoning.

This work directly inspired linked.archi's own principle-based validation shapes, which implement 20 SHACL shapes for architecture governance principles including redundancy avoidance, single source of truth, stewardship, separation of concerns, service orientation, completeness, and motivation alignment.

Key Insights from This Work

  • Academic rigor strengthens ontology design. Competency questions, practitioner interviews, and multi-project validation provide a level of confidence that informal development cannot match. The methodology described in the MODELSWARD 2020 paper is a reference for anyone building enterprise ontologies.

  • Enterprise domain content has value. Organizational roles, document types, business events, and geographic reference data — while mixing concerns from a pure ontology engineering perspective — make the ontology immediately usable for real enterprise applications without requiring users to build their own domain vocabulary from scratch.

  • Cross-notation integration via class hierarchy is powerful. Making BPMN elements subclasses of ArchiMate concepts enables natural cross-notation reasoning. The trade-off (tight coupling) is acceptable when the goal is a unified enterprise ontology rather than a modular toolkit.

  • ISO 42010 formalization is underexplored. ArchiMEO's dedicated ISO 42010 module — particularly Correspondence and CorrespondenceRule — addresses governance concepts that most ArchiMate formalizations skip. These map naturally to SHACL shapes for cross-view consistency validation.

  • BMM bridges motivation to architecture. The explicit Business Motivation Model module provides concepts (Vision, Mission, Strategy, Tactic, BusinessPolicy, BusinessRule) that complement ArchiMate's Motivation layer and are valuable for organizations using both standards.

  • Continued evolution matters. The 2026 principle validation paper shows ArchiMEO actively adopting modern standards (SHACL) while building on its established foundation — a healthy pattern for any long-lived ontology project.

  • Semantic lifting is a shared concern. ArchiMEO's demonstrated 1:1 correspondence between graphical models and ontology instances validates the same approach that linked.archi implements through its converter toolchain (ArchiMate Exchange XML → RDF, BPMN 2.0.2 XML → RDF).

References

  1. Hinkelmann, K., Laurenzi, E., Martin, A., Montecchiari, D., Spahic, M. and Thönssen, B. (2020). ArchiMEO: A Standardized Enterprise Ontology based on the ArchiMate Conceptual Model. In Proceedings of MODELSWARD 2020, pp. 417–424. DOI: 10.5220/0009000204170424.
  2. Hinkelmann, K., Gerber, A., Karagiannis, D., Thoenssen, B., Van der Merwe, A., and Woitsch, R. (2016a). A new paradigm for the continuous alignment of business and IT. Computers in Industry, 79:77–86.
  3. Hinkelmann, K., Laurenzi, E., Martin, A., and Thönssen, B. (2018). Ontology-based metamodeling. In Business Information Systems and Technology 4.0, pp. 177–194.
  4. Martin, A. (2016). A combined Case-Based Reasoning and Process Execution Approach for Knowledge-Intensive Work. PhD thesis, University of South Africa.
  5. Hinkelmann, K. et al. (2026). Ontology-Based Validation of Enterprise Architecture Principles. Applied Sciences, 16(7):3352.
  6. Linked.Archi ArchiMate Design Decisions
  7. Linked.Archi Relationship Modeling Guide
  8. Linked.Archi BPMN Deep Dive
  9. Mendoza ArchiMate Ontology
  10. bp4mc2 archimate2rdf
  11. Figay ArchiMate OWL