SWAT4HCLS 2026 Biohackathon

Overview

Semantic Web Applications and Tools for Health Care and Life Sciences (SWAT4HCLS) is an annual conference at the intersection of semantic web technologies and biomedical research. Each year, the conference includes a dedicated Biohackathon track, a collaborative working session that runs in parallel with the main conference and continues with a dedicated full day afterwards.

I participated in the SWAT4HCLS 2026 Biohackathon in Amsterdam, Netherlands, leading a team looking at how national healthcare metadata catalogs across Europe describe their data, and what it would take to make them discoverable and queryable as a federated whole.

Projects

From DCAT to HealthDCAT-AP: Comparative Analysis of National Healthcare Metadata Catalog Schemas to Enable Federated Interoperability in Europe

Across Europe, health-related data catalogs are converging on DCAT-AP and its healthcare-specific extension, HealthDCAT-AP, as the common vocabulary for describing and publishing datasets. In practice, countries and institutions tend to adapt these base schemas to their own regulatory, organizational, and technical needs. The result is a patchwork of national profiles that are individually well-formed but diverge from one another in subtle ways (e.g., different mandatory fields, different controlled vocabularies, different ways of encoding the same underlying concept) which makes it difficult to discover and reuse health datasets across borders.

The main objective of this project was to map that landscape systematically: to compare the underlying schemas of health-related metadata catalogs from several European countries against DCAT-AP and HealthDCAT-AP, and against each other. The team went field by field through each national implementation, cataloging where definitions lined up cleanly, where they diverged in scope or cardinality, and where a concept present in one schema was simply absent from another. This produced a concrete picture of which parts of the ecosystem are already interoperable by default and which parts would need explicit crosswalks or schema changes to support federated search.

The work addressed challenges in discoverability and interoperability of health-related datasets caused by divergent metadata standards and national adaptations, with a particular focus on the requirements set out by the European Health Data Space (EHDS), which aims to let researchers and health systems query and combine datasets across EU member states.