Interfacing CLARIN with H2IOSC: Metadata Interoperability through Ontology-based Mediation

Abstract

We present an ontology-based mediation approach for integrating CLARIN language resources into the H2IOSC semantic framework, with the aim of enabling semantic interoperability across the Social Sciences and Humanities. Building on CMDI and CLARIN CCR, our work develops a layered semantic alignment workflow in which CLARIN metadata are progressively reinterpreted through multiple mediation steps. In a first layer, CMDI metadata are aligned with CIDOC CRM and SSHOCro, which are extended through a CLARIN domain ontology that formalises distinctions between datasets, tools, and services as reflected in discovery environments such as the Virtual Language Observatory, Resource Families, and the Language Resource Switchboard. In a second layer, these domain-level representations are projected into H2IOSCro, a Marketplaceoriented ontology that introduces entities, attributes, and typed relations required for operational federation within H2IOSC. The resulting ontology-based representation is finally projected onto the Marketplace operational data model, supporting metadata harvesting, normalisation, validation, and exposure through standard mechanisms such as OAI-PMH. By explicitly connecting semantic modelling, ontology mediation, and platform-level integration, this layered approach enhances discoverability, preserves compatibility with existing CLARIN metadata practices, and contributes to FAIR- and EOSC-aligned interoperability.