We present the Data Set Knowledge Graph (DSKG.org), an RDF dataset about datasets that are linked to publications (modeled in the Microsoft Academic Knowledge Graph, MAKG) that mention the datasets. The metadata of the datasets is based on datasets that are registered in OpenAIRE and Wikidata.
What exactly do we provide?
- Periodically updated RDF dump files of the Data Set Knowledge Graph.
- URI resolution of the Data Set Knowledge Graph within the Linked Open Data.
- A publicly accessible SPARQL endpoint containing the latest Dataset Knowledge Graph data.
How big is the Dataset Knowledge Graph?
The Dataset Knowledge Graph models, among others,
- 2,208 datasets from all scientific disciplines
- 813,551 links to 634,803 unique papers
- 1,169 authors of datasets
- 208 ORCID IDs.
Potential use cases:
- Use the DSKG for the development of semantic search engines (e.g. use the metadata of the linked publications of the datasets for advanced search capabilities)
- Easier data integration by using the RDF standard vocabulary DCAT and by linking resources to other data sources (e.g., combining the DSKG with other dataset collections in RDF).
- Data analysis to measure and award the provisioning of datasets (e.g., determine the scientific influence of datasets and authors).