Federated Querying

On this page

    One of the key features of Comunica, is the ability to query over multiple sources of different types. This concept of querying over multiple sources is called federated querying.

    This functionality can be exploited on both the CLI and the JavaScript API. In this guide, we will make use of the CLI as an example.

    Federated query execution does not just send the query to each source separately. Instead, the triples from all sources are considered one large virtual dataset, which can then be queried over.

    Distributed Knowledge

    A fundamental concept of Linked Data and the Semantic Web is that data can be spread over different sources across the Web. This means that querying over this data potentially involves more than one source.

    While some knowledge graphs such as DBpedia and Wikidata aim to accumulate as much data as possible in one place, these always have limitations in scope. As such, federated querying may be needed for some queries.

    Federated Querying in Comunica

    Comunica's ability to execute federated queries is enabled by default. This can be invoked by simply passing more than one source to the engine.

    For example, the following query will retrieve all triples from DBpedia and two RDF documents:

    $ comunica-sparql https://fragments.dbpedia.org/2016-04/en \
        https://www.rubensworks.net/ \
        https://ruben.verborgh.org/profile/ \
        "SELECT * WHERE { ?s ?p ?o }"

    The example above shows that sources do not necessarily have to be of the same type.

    Real-world federation example

    One example of a real-world federated query, is task of linking people in DBpedia to library datasets. For this, the Virtual International Authority File can be used as a source to provide this linking.

    The query below will retrieve all books in the Harvard Library written by people born in San Francisco:

    $ comunica-sparql https://fragments.dbpedia.org/2016-04/en \
        http://data.linkeddatafragments.org/viaf \
        http://data.linkeddatafragments.org/harvard \
        'SELECT ?person ?name ?book ?title {
           ?person dbpedia-owl:birthPlace [ rdfs:label "San Francisco"@en ].
           ?viafID schema:sameAs ?person;
                        schema:name ?name.
           ?book dc:contributor [ foaf:name ?name ];
                       dc:title ?title.
    The TPF-based source https://fragments.dbpedia.org/2016-04/en is interchangeable with SPARQL-endpoint-based source https://dbpedia.org/sparql. The engine will produce similar results as the sources represent the same dataset.