“Moselle wine, LLMs, and LOD – analyzing cultural heritage with AI”

The Semantic Web and Linked Open Data is a long-established framework for data processing that has recently become increasingly relevant. This presentation explains the basics of this framework, what Wikibase and Wikidata actually are, and then shows a concrete application example from a current research project on local cultural heritage. Using the example of wines from the Moselle region and their labels, it shows how to create ontologies and vocabularies for an LOD project. It also reports on how modern multimodal large language models can be used to populate an LOD database for image elements. These models rely on a combination of image and text data that support each other in inference.