News for hackers. Stuff that matters.
Re:Building New SQL (2013) [pdf] by jerven on Tuesday September 02, 2014 @09:13AM (#8257300)
The SPARQL/RDF world is moving to that "figure out the best storage solution itself". You just throw data into a system and the system figures out how best to store it. There is some research code for this in MonetDB and the openlinksw/virtuoso guys are hinting about it in their latest blog posts for their upcoming release. Basically as some people like to call it, they want to infer the relational schema hiding in your RDF data.

Currently most SPARQL endpoint store data in two tables a quad/triple table and a big id to value table. But there is no need to do so.

Main problem with SPARQL and RDF problem is that as the schema is in the data its difficult to figure out how data is connected and how one can use it. But the W3C is starting up a workgroup to work on that problem, for tool builders to solve this.

There are no comments on this comment yet.