Semantic Uplift
Obstacle
Implementing the ambitious strategies of the Green Deal require connecting a multitude of heterogeneous services and datasets. Since a single overall solution is not possible, interoperability among components of the GDDS is required, especially at a semantic level.
Solution
AD4GD has enabled semantic interoperability, harmonisation, and enrichment across all the components developed for the GDDS. The approach includes reusing existing and widely utilized ontologies under the framework of the Green Deal Information Model (GDIM) and the establishment of semantic mappings that bridge domain- and source-specific semantic concepts, ensuring reliability and trustworthiness of the data.
Usage in the project
Semantic uplift was applied in all three AD4GD pilots to harmonise local datasets with the GDIM. Each pilot implemented harmonisation semantic pipelines, leveraging shared vocabularies and mapping specifications. In particular, the biodiversity pilot used SPARQL mappings and EV vocabularies to semantically transform species simulation data (e.g., BES-SIM AIM) into interoperable RDF representations. However, the water and air quality pilots employed YAML-based mappings and custom vocabularies processed via DPI pipelines and published through the OGC Rainbow server, ensuring semantic traceability and alignment with European standards.
Foreseen applications
The semantic uplift services and tools developed by AD4GD can be exploited for the development of the GDDS. All the functionalities can be replicated and reused in other environments according to the user needs.
Component led by:
Featured pilots
Learn more
How does AD4GD's semantic uplift work?
The Semantic Uplift component is central to enabling interoperability, harmonisation, and semantic alignment across AD4GD ecosystem. It provides a suite of modular, cloud-hosted semantic services designed to transform heterogeneous datasets into semantically enriched and interoperable resources.
These services support structured alignment with ontologies such as the Green Deal Information Model (GDIM), which is core to the GDDS. Semantic uplift lower entry barriers to streamlined tooling and automation, empowering both technical and non-technical stakeholders to contribute data with minimal configuration, typically requiring only initial mapping to shared vocabularies.
The Semantic Uplift component delivers a comprehensive set of capabilities through integrated services designed to support semantic interoperability, data harmonisation, and compliance with EU data frameworks. These capabilities are realised through the following key functionalities and tools:
Data Mapping, Harmonisation & Semantic Lifting
It transforms heterogeneous datasets into semantically enriched, interoperable formats aligned with the GDIM and its underlying re-used standard models like SOSA/SSN, QUDT, and other ontologies. Based on the use of Linked Data and Knowledge graph technologies, the tool facilitates data harmonisation pipelines, CLI, Web Service, and GUI-based tools for ETL, semantic mapping, transformation, and linking.
Vocabulary Alignment & Semantic Publishing
It publishes vocabularies that align datasets related concepts (e.g., observed properties or sensors) with standard vocabularies and thesauri (e.g. Climate and Forecast standard names, Agrovoc, Essential Variables) to ensure semantic coherence across domains. The tool facilitating this is the OGC Rainbow Server, a public registry for publishing and browsing domain-specific vocabularies, mappings, and identifiers with dereferencable URIs.
Harmonized data access components
The semantic uplift is able to provide or generate interfaces to access the GDIM harmonized data. This is achieved through access interfaces, such as SPARQL and OGC APIs, and components to generate standard APIs, including the OGC STA generation service that generates STA endpoints automatically from GDIM compliant measurements or observations.
Trusted Data Processing Environments
Cloud-hosted environments are provided to execute semantic uplift workflows, including annotation and transformation pipelines, using DPI Cloud Services. These, support CSVW, JSON Schema, SHACL validation, and storage in semantic triple stores such as Virtuoso.