Semantic Data Specifications must serve the requirements of business experts and technical developers while ensuring all representations remain consistent and clear for accurate machine interpretation.
Two foundational concepts address these challenges: the separation of concerns and the Single Source of Truth (SSoT). These principles allow Knowledge Engineers to automatically transform conceptual models into precise SDS artefacts. This approach minimizes errors, streamlines maintenance, and ensures that every artefact remains consistent with the core model.
In this article, we’ll delve into how SEMIC’s principles and model2owl enable the creation of reliable, automated, and truly interoperable Semantic Data Specifications.
1. Separation of Concerns in Semantic Data Specifications
Creating an effective Semantic Data Specification (SDS) is more than just building a robust ontology. It requires integration into a development and implementation process that addresses both domain-specific and technical needs.
🟢 The separation of concerns implies the separation of representations.
By isolating different representations, SDS can address distinct needs across audiences and lifecycle stages, ensuring clarity and usability for all stakeholders.
Each model’s utility depends on which concerns it addresses, as underscored by the statistician George Box’s insight: “All models are wrong, but some are useful.” In SDS, models must simultaneously represent the conceptual domain and technical system requirements. For example, “location” may have a broad meaning for domain experts but require specific attributes in technical applications. To manage this, SEMIC uses the separation of concerns to isolate domain knowledge from implementation details while maintaining consistency.
🟢 Model-Driven Architecture
SEMIC adopts the Model Driven Architecture (MDA) framework, which creates a model for each concern and then transforms it to meet other objectives. This approach allows SDS to balance conceptual and technical needs, adapting models through transformation patterns. Each transformed model provides a different view, whether it’s for business requirements, technical constraints, or interoperability standards.
As defined here, transformation is the process of creating various models or artefacts from a central one, following specific guidelines. These transformations can span levels of abstraction, from conceptual overviews to detailed specifications. In SEMIC, these transformations allow the SDS to serve the diverse requirements of all stakeholders.
🟢 Applying Separation of Concerns in SEMIC
SEMIC implements this approach through Core Vocabularies (CVs) and Application Profiles (APs), each focusing on distinct needs:
→ Domain Conceptualization
Domain experts need to agree on definitions, relationships, and organization within a shared domain model, expressed in UML (Unified Modeling Language) to facilitate clarity for non-technical stakeholders.
→ Machine Usability
Machines need a format they can interpret. Lightweight ontologies in OWL 2 (Web Ontology Language) encode these models to enable interoperability across systems.
→ Interoperability Through Data Constraints
Interoperability requires specific constraints on the ontology provided by SHACL (Shapes Constraint Language). This set of rules ensures consistency across applications while preserving flexibility.
→ Clear Documentation
To foster understanding and adoption, HTML documentation provides a human-readable format accessible to all stakeholders, including developers and domain experts.
These distinct representations allow the SDS to be both flexible and consistent, supporting a wide range of use cases.
2. The Solution: Editorial Synchronization and The Single Source of Truth (SSoT)
While separating concerns provides clarity, it also introduces the challenge of synchronizing representations. Changes to one artefact (be it UML, OWL, or SHACL) must propagate consistently to maintain alignment. Manually synchronizing these representations can be labor-intensive and error-prone, compromising the SDS’s coherence.
SEMIC uses a Single Source of Truth (SSoT) approach to address this, where a central model (often the UML conceptual model) serves as the definitive version. From this single model, all other representations can be derived automatically, streamlining updates and ensuring consistent alignment across artefacts.
Through this separation of concerns and structured transformation, SEMIC provides a framework that enables Knowledge Engineers and domain experts to create SDS that are both technically robust and conceptually sound, supporting true interoperability:
🟢 The Core Model as the Truth
Instead of creating each artefact independently, the SSoT uses the conceptual model as a unified source. This model is expressed in UML, which captures both domain-level and technical concerns. Different representations can be derived from this single model as needed, ensuring that every artefact (from visual diagrams to data shapes and ontologies) originates from the same consistent base.
🟢 Deriving Artefacts from SSoT
By using the conceptual model as the foundation, artefacts can be generated for different needs without additional manual steps. This approach differs from many ontology development methodologies, where each representation often requires manual intervention for encoding and documentation. With an SSoT, transformations are automated, leading to fewer errors and better consistency across the entire SDS. Changes made to the conceptual model automatically propagate to other artefacts, reducing the burden on editors.
🟢 Editorial Synchronization
The SSoT addresses the common synchronization problem by grounding all artefacts in a single model. This efficient approach reduces the risk of discrepancies between representations, keeping the SDS cohesive and aligned with both business and technical needs.
3. Automatic Transformation of Single Source of Truth (SSoT) into SDS Artefacts using model2owl
To create robust semantic data specifications (SDS), organizations will find that relying on a Single Source of Truth (SSoT) is no longer optional. The SSoT model addresses the complexities and variances between technical and business requirements by acting as a unifying, authoritative base from which every SDS artefact is derived.
But transforming this SSoT into multiple, compatible SDS artefacts that are machine-readable and human-interpretable is far from simple.
Therefore, tools like model2owl become a necessary asset.
🟢 Transforming Conceptual Models into SDS Artefacts with model2owl
model2owl directly supports the SEMIC architecture by automating the conversion of a single conceptual model into various SDS artefacts, particularly OWL (Web Ontology Language) ontologies and SHACL (Shapes Constraint Language) data shapes.
Through this automation, model2owl ensures that every artefact produced is aligned with the conceptual model, reducing the margin for error typically associated with manual updates. As the core model evolves, model2owl allows changes to ripple across artefacts seamlessly, maintaining consistency without additional editorial effort.
The tool’s functionality is rooted in a structured transformation pattern. model2owl takes the conceptual model, often represented in UML, and systematically derives specific outputs for different user contexts. OWL ontologies are generated to encode domain knowledge in a format that is accessible for machine processing, while SHACL data shapes ensure that the data structure adheres to defined constraints, fostering interoperability. This transformation aligns perfectly with SEMIC’s vision by grounding technical artefacts in the same concepts and relationships agreed upon by domain experts.
🟢 Preserving the integrity of the Single Source of Truth across representations.
model2owl ensures the integrity of the Single Source of Truth across representations. Editors no longer need to manually adjust each artefact, mitigating the risk of inconsistencies and versioning conflicts. Moreover, the alignment of business concepts and technical models is preserved, as every artefact speaks to the same foundational principles captured within the SSoT.
🟢 A Solution for Knowledge Engineers
For knowledge engineers, model2owl brings considerable efficiency, translating the complexity of multi-representational needs into an automated process that upholds the SEMIC principle of “separation of concerns.” Instead of creating isolated artefacts that require ongoing manual synchronization, model2owl outputs artefacts that naturally adhere to the SSoT, reinforcing coherence and usability across the SDS.
model2owl exemplifies how SEMIC’s architecture for SDS can be applied practically. This tool facilitates automatic transformations and enables Knowledge Engineers to maintain a high standard of consistency and reliability across every SDS artefact, establishing a resilient framework for cross-functional interoperability.
4. We Believe in a Future of Fully Interoperable Semantic Data Standards
As organizations move toward data-driven decision-making and cross-border collaboration, the role of a well-structured Semantic Data Specification (SDS) cannot be overlooked. Through the strategic application of the Single Source of Truth (SSoT) and tools like model2owl, Knowledge Engineers are now better equipped to maintain alignment across business, technical, and machine contexts. This alignment supports unprecedented clarity and interoperability, precisely unifying abstract domain knowledge and technical implementation.
The future of SDS lies in advancing automation and maintaining coherence across every level of data representation. SEMIC’s principles of separation of concerns and transformative tools signal a move toward an interconnected, standards-based ecosystem.
We invite you to explore this article for additional insights about Semantic Data Specifications.
Please visit this article for more information about the Single Source of Truth.
You can check more information about model2owl in this article.
Meaningfy continues to support the European Commission’s initiatives, leading the charge toward a transparent, efficient, and interconnected European public sector. If you represent a European Institution or a public company that needs to implement an interoperability solution, contact us, and we’ll help you implement it effectively.