Publishing LOD Cloud SPARQL Endpoint Descriptions using Linked Data Principles

The LOD Cloud Knowledge Graph spawned by DBpedia is the world’s largest public Knowledge Graph, by far! It is also a practical demonstration of the vision that underlies the notion of a Semantic Web constructed from hyperlinks based on the following principles:

  1. Identify (or Name) anything unambiguously using a hyperlink (specifically, a HTTP URI)
  2. Describe everything using RDF structured sentences where a hyperlink identifies the subject and predicate, and optionally the object (which may also be a typed or untyped literal)
  3. Save your RDF-based sentences to a document
  4. Publish your document to the World Wide Web

The net effect of this “deceptively simple” approach to creating and sharing information is that it provides a powerful Knowledge Graph collective on the Web (also known as the Linked Open Data Cloud) that manifests as a fine-grained Web of Data rather than a coarse-grained Web of Documents. Consequently, whenever you click on a hyperlink it will return a document that describes whatever the hyperlink identifies; in other words, you have a property sheet for every entity (every thing) identified by a hyperlink.

Once you have a Web of Data in place, human interaction (via software agents) with the knowledge it exposes is a natural next step, which is one area where the SPARQL Query language, protocol, and query solution document types offer immense value.

The LOD Cloud Knowledge Graph and associated SPARQL Query Service Endpoints provide data access infrastructure similar to an electric grid where hyperlinks function as conductors of data-flow across a federation of data sources. Thus, describing these endpoints using RDF sentences and then deploying said descriptions using Linked Data principles enables simplified discovery and exploitation by end-users, data integrators, and application developers.

Courtesy of our Virtuoso RDBMS – the dominant SPARQL Query Service provider in the LOD Cloud Knowledge Graph – here is a collection of links, associated with the description of currently discovered and active SPARQL Endpoints, that has been deployed using Linked Data principles.