Home

SPARQLFederator - Expand SPARQL queries to perform inference on multiple endpoints and relational databases.

SPARQLFederator is a tool that rewrite SPARQL queries to achieve reasoning over:

• data distributed on several endpoints;
• data on an endpoint without reasoning capabilities.

SPARQLFederator is released under the GNU/GPL v3 license.

Official repository

Mailing list

Use cases

Contribute

How to cite

What does it do?

The possibility of combining open data from several sources is the great advantage of the semantic web. Since the publication of SPARQL 1.1 standard, several semantic web tools provided this funcionality. However, the specification of sources has to be done manually. Worse, if the user desires to perform inference on all data from several different sources, a complicated pipeline has to be set up.

SPARQLFederator approaches those problems by query expansion:

• For each given triple, all applicable endpoints are queried; its results are UNIONed.
• When a queried class is part of an OWL axiom, it is expanded in its component parts.

SPARQL Federator may also be useful when inference is needed on a single endpoint without reasoning capabilities.

How it is done?

SPARQL Federator needs three things to get its job done:

• An OWL ontology telling where are the endpoints, and which properties and classes they have data about (the federation ontology)
• An OWL ontology containing axioms about the classes (the domain ontology)
• A SPARQL query

It may return the expanded query, or run it (using the ARQ-Jena Library).