I’ve been looking for a SPARQL endpoint that provides new data fairly regularly—not just new triples to query, but data that is new to the world, such as from a stock ticker feed. If the RDFa on digg.com pages was accumulated in a database that could be queried as a SPARQL endpoint, that would certainly qualify, and it would be fun to play with.
http://demo.openlinksw.com/sparql or /isparql
This instance of Virtuoso includes are in-built Sponger Middleware.
The Sponger Middleware converts a plethora of none RDF resources into RDF based Linked Data “on the fly”.
The Sponger is also integrated into the Virtuoso SPARQL processor so you can put any resource URL in the “FROM” clause of a SPARQL query. The effect is that you can SPARQL against any Web resource URI via a Virtuoso sparql endpoint.
1. If a local graph IRI matching the resource URL doesn’t exist, the Sponger will crawl the resource
2. The localized resource is then RDFized (we have RDFizers aka. Cartridges for about 30 different data source types which includes Digg)
3. The Graph IRI for the sponged resource is always the same as the original resource URL.
Basically, the Sponger is like a Driver Manager, but instead of dealing with relational data (ala. ODBC, JDBC etc..) it offers dynamic binding to RDF Drivers / Providers / Cartridges which take on the duty of transforming negotiated resource representations into RDF based Linked Data.
I put up an endpoint for one of my projects. It’s not exciting data, but it is new each day.