1. Introduction

The NFDI4Culture communities, including architecture, art history, design, musicology, media studies, the performing arts, digital humanities, and contributors from the creative sector and cultural heritage, are encouraged to provide metadata about their research data for inclusion in the Culture Knowledge Graph (PID E2391). Metadata here refers to information about research data held by data providers in NFDI4Culture’s data domains.

This guideline is designed for data providers to understand the purpose of the Culture Knowledge Graph as an overarching metadata index (rather than a repository for data storage). It outlines the supported data formats as well as provisioning methods and provides detailed guidance on the specific considerations and steps required for data integration into the Culture Knowledge Graph.

Please ensure that you consult the newest version of this guideline as chapters will regularly be revised or updated. Additional extraction/transformation routines for widely used community standards are in development, and further transformation routines developed by the community, even for specific data providers, are welcome. A list of abbreviations and a glossary are available in the appendix.

Ten things to know

1. The Culture Knowledge Graph is designed to interconnect research data through metadata and make it more findable.
2. It represents your data via a limited set of metadata properties to make it available across all of NFDI4Culture’s research domains as well as to other NFDI consortia.
3. The ontology used in the graph (i.e. NFDIcore/CTO) is designed to guide users towards more detailed research data on the data providers’ websites.
4. Hence, resolvable IRIs, such as permalinks or persistent identifiers (PIDs), are a key requirement of getting your data into the graph.
5. Connections across the graph highly benefit from the use of authority data to increase interoperability.
6. Metadata about research data from the cultural domains is provided via data feeds and their data feed elements.
7. The integration of data feeds is managed via a flexible Extract-Transform-Load (ETL) environment, which allows reharvesting in case of modifications.
8. A data feed needs to be registered in the Culture Information Portal (PID E2379) before it can be ingested.
9. A SPARQL endpoint for the Culture Knowledge Graph is already operational, and data providers can query it to, for example, show related content to their users.
10. A visual search and exploration interface is currently under development.