How to download catalog data

This page explains the supported ways to download one KG, one version, or the full catalog from Databus and related sources.

Download the latest version of a KG

First, open the KG profile page (e.g., ORKG) and identify the Databus identifier for the knowledge graph you want to download.

Then, use the following Docker command to download the latest version of the KG. The example below is for ORKG:

docker run --rm \
  -v "$(pwd)":/data \
  dbpedia/databus-python-client \
  download https://databus.dbpedia.org/knowledge-graph-catalog/orkg

Download all versions of a KG

First, open the KG profile page (E.g., ORKG) and identify the Databus identifier for the knowledge graph you want to download.

Then, use the following Docker command to download all versions of the KG. The example below is for ORKG and uses the --all-versions flag:

docker run --rm \
  -v "$(pwd)":/data \
  dbpedia/databus-python-client \
  download --all-versions https://databus.dbpedia.org/knowledge-graph-catalog/orkg

Download the latest versions of all KGs

Download the latest version of each artifact for every KG in the catalog with one command. The script discovers KGs through Databus and downloads up to 3 KGs in parallel by default.

curl -fsSLO https://raw.githubusercontent.com/dbpedia/kg-catalog/main/scripts/download-all-kgs.sh && bash download-all-kgs.sh

To change the number of simultaneous KG downloads (e.g., to 10), run:

PARALLEL_DOWNLOADS=10 bash download-all-kgs.sh

Other configuration options

Set these environment variables before bash download-all-kgs.sh:

MAX_SECONDS — timeout per KG in seconds (default: 172800, 48 hours; 0 disables).
MAX_KGS — limit how many KGs are downloaded (default: 0, all KGs).
OUTPUT_DIR — directory for downloads and reports (default: kg-catalog-download-test beside the script).
DISCOVERY_TIMEOUT — discovery request timeout in seconds (default: 45).
GROUPS_FILE — optional text file with one Databus group URL per line, to download selected KGs.
USER_AGENT — HTTP User-Agent to use for downloads (default: DBpedia Databus Python Client).
VENV_DIR — optional location for the local Python virtual environment.

For example: MAX_KGS=5 MAX_SECONDS=3600 bash download-all-kgs.sh

The script sets up its Python dependencies automatically and saves downloads and a per-KG report locally. Some files require a Vault token (vault-token.dat alongside the script). Large KGs can require substantial disk space and download time. Review the script on GitHub before running it.