H

Helmholtz-Knowledge-Graph

Manages JSON LD data being injected into, and data dumps coming from Virtuoso Graph DB

This is the internal data model used in the Helmholtz Knowledge Graph Infrastructure. RAW metadata that was harvested is mapped on this data model for harmonisation and validation purposes and to supply data to the UI.

Set of tools to harvest, process and uplift (meta)data from metadata providers within the Helmholtz association to be included in the Helmholtz Knowledge Graph (Helmholtz-KG). The harvested linked data in the form of schema.org jsonld is aggregated and uplifted in data pipelines to be included into a single large knowledge graph (KG). The tool set and harvesters can be used as a python library or over a commandline interface (CLI, hmc-unhide). Provenance of metadata changes is tracked rudimentary by saving graph patches of changes on rdflib Graph data structures on the semantic triple level. Harvesters support extracting data via sitemap, gitlab API, datacite API and OAI-PMH endpoints.

This repository contains docker compose deployment files for the whole software stack behind the Helmholtz knowledge graph, work from the unHIDE initiative. Deployment includes containers for the harvesters and utility, the API, SOLR, Virtuoso, Web Frontend as well as for nginx and letsencrypt. More information on the unHIDE initiative, which was launched by the Helmholtz Metadata Collaboration (HMC), you can find under https://docs.unhide.helmholtz-metadaten.de.