Data · dataset · 2026
D7.1 A Framework for Data, Domain, and Process Modelling in Cross-Domain Harmonization
Listed in D2ET Open Science Portal
This deliverable presents the methodological framework and initial results underpinning the data and model harmonization activities of the work package.
Description
It addresses the challenge of integrating heterogeneous data sources, domain knowledge, and analytical components across a multi-domain decision-making platform by proposing three complementary and mutually reinforcing modelling instruments: a data cartography and initial alignment framework, a conceptual modelling approach, and a functional modelling methodology.
Together, these instruments constitute a layered knowledge structure designed to support semantic coherence, interoperability, composability, and explainability across the platform. The deliverable first introduces a methodological framework for domain modelling inspired by Domain‑Driven Design principles, that combines a bottom‑up perspective, in which existing datasets are analyzed to derive consistent data models, with a complementary top‑down perspective, in which domain concepts and structures guide the organization of the data and the design of the platform components.
Read the rest (6 more)
A motivation modelling layer, grounded in the ArchiMate Motivation extension, provides the strategic and semantic context within which all three instruments operate. A key component of the work is the development of a data cartography, supported by WP9 (Task 9.1), which maps the available data resources and their relationships. This cartography extends beyond traditional data models by incorporating metadata that describe the context of the data, including data sources, ownership, temporal and spatial granularity, quality indicators, and provenance.
A dual-layer provenance model (combining a DCAT-AP and PROV-O conceptual representation with an operational OpenLineage lineage capture mechanism) supports traceability and reproducibility across the platform's data pipelines. The conceptual modelling approach addresses the semantic dimension of the platform: how the key concepts used across five distinct contributing domains are defined, related, and made explicit to support coherent integration.
A purpose-built, lightweight notation language (structured around directed relationships, typed property and derivation relationships, and explicit concept definitions) enables domain experts to construct comparable concept maps without requiring formal ontology engineering skills. A systematic cross-domain semantic analysis, currently under way, will derive a shared platform vocabulary from the comparison and reconciliation of these maps.
The functional modelling methodology operationalizes a model-based engineering approach through a standardized functions template. This template captures, for each platform component, its purpose, capabilities, valid operating conditions, modelling formalism, algorithms, execution requirements, and interface specification. The template is designed to be interpretable by both human collaborators and AI agents, providing the structured knowledge base required to support automated composability assessment, semantic interoperability checking, and explainability of composed results.
The template has been distributed to and completed by contributing teams from WP1 to WP4, with an initial cross-analysis revealing semantic inconsistencies across domains that confirm thecore motivation of the methodology. The deliverable concludes that the three modelling instruments are most valuable when developed simultaneously and in relation to one another. The data cartography grounds the functional model in concrete data assets; the conceptual model assigns unambiguous meaning to the concepts that circulate across the platform; and the functional model captures the operational identity of each component in a form consistent with that shared vocabulary.
This layered knowledge structure constitutes thefoundational element of the Noetic Layer of the D2ET platform, enabling coherent integration of data, models, and analytical services across the project ecosystem.
Links
Where it is published
- Dataverse dataset page d2et-openscience.list.lu/dataset.xhtml?persistentId=perma%3AD2ET.9ITTQF ↗
landing page · from d2et openscience list lu
- Persistent identifier d2et-openscience.list.lu/citation?persistentId=perma%3AD2ET.9ITTQF ↗
DOI / persistent id · from d2et openscience list lu
Catalogue records · 1
- Dataverse API d2et-openscience.list.lu/api/datasets/:persistentId/?persistentId=perma%3AD2ET.9ITTQF ↗
metadata API · from d2et openscience list lu
Topics
- Stated by source
- Computer and Information Science
- Inferred from text
- Information modelling, management and ontologies 79%
Provenance · 1 source records, 8 field assertions
| Source | Key | Last seen | Raw |
|---|---|---|---|
| D2ET Open Science Portal | perma:D2ET.9ITTQF | 5 d ago | JSON v1 |
| Field | Assertion | Extractor | Evidence |
|---|---|---|---|
| concepts[field].anzsrc:field:460903 | enrichment · d2et openscience list lu | taxonomy-embedding@1.0.0 | title+keywords+description (79%) |
| concepts[field].dataverse_subject:computer-and-information-science | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | /subjects |
| created_date | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | |
| description | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | /description |
| publication_date | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | |
| title | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | /name |
| updated_date | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 | |
| version_label | source · d2et openscience list lu | connector:d2et_openscience_list_lu@1.0.0 |