DBpedia KG in Neo4J

This repository provides simple utilities to populate a Neo4J graph database instance with a certain version of the DBpedia KG using the neosemantics (n10s) plugin. Loading the complete DBpedia KG into Neo4J might take hours depending on your hardware, however.

MATCH (x:DBI) -[p]-> (y:DBI)
WHERE x.uri = "http://dbpedia.org/resource/Capybara"
RETURN x, p, y

All nodes are labelled with Resource by neosemantics. Additionally, all nodes that have a URI/IRI starting with http://dbpedia.org/ are labelled with DBI.

Download Dump

The DBpedia Core (latest core) collection can be downloaded from the DBpedia databus with the following script. If no argument is specified, then the latest dump is downloaded, otherwise the specified dump date (e.g. 2022-03-01).

The requirements for the script are:

• bzip2
• gzip
• wget & curl

./get-dbpedia-dump [<dump-version>]

All artifacts are downloaded from the latest-core collection, except for:

• all artifacts from the replaced-iris dataset

Additionally to the latest core collection, the following artifacts are added to the collection:

• wikilinks from generic dataset
• images from generic dataset

The exact SPARQL query for gathering can be found at query/data.query.

Running

docker-compose up

Contact

• Kevin Haller - [email protected]