08/07/2026
On June 18, 2026, Thora Hagen (IDS Mannheim) delivered the first DataGEMS Language Pilot Study Workshop, hosted by the Leibniz Institute for the German Language, Department of Digital Linguistics (online). The session presented DataGEMS, a next-generation platform for discovering, analyzing, and reusing open data across domains, developed within the EOSC ecosystem.
The workshop focused on the Language Use Case, which addresses the structured exploration, thematic search, semantic analysis, and educational transformation of multilingual text corpora through five sequential pilots:
- Corpus Profiling (UC4P1). The currently active pilot, offering a feature-based representation of unstructured texts, using two editions of the Encyclopædia Britannica as its data basis.
- Thematic Search (UC4P2). Domain-specific knowledge synthesis from corpora, with a focus on content.
- Semantic Change (UC4P3). Diachronic analysis of meaning change, with a lexical focus.
- Educational Transformation (UC4P4). Simplification of complex content for learners.
- Human-in-the-Loop (UC4P5). Expert guidance and pipeline control throughout all pilots.
The presentation also addressed why existing systems are not optimal, limited search, poor metadata, and single data type support, and how DataGEMS overcomes these through content-aware search across datasets, integration of different modalities, and transparent use of modern LLMs, all in line with FAIR principles.
This workshop represents an important milestone for the DataGEMS project, demonstrating significant progress toward FAIR, content-aware data discovery and exploration for the digital humanities and computational linguistics communities.
