W3C home > Mailing lists > Public > semantic-web@w3.org > May 2014

Re: Named Graphs

From: Simon Spero <sesuncedu@gmail.com>
Date: Thu, 15 May 2014 19:35:45 -0400
Message-ID: <CADE8KM7mLBCRPxLjOczP3i8NsA7w3b6e3BCfA8QCXAM0sZoOiA@mail.gmail.com>
To: Jeen Broekstra <jeen.broekstra@gmail.com>
Cc: "semantic-web@w3.org" <semantic-web@w3.org>
On Thu, May 15, 2014 at 6:30 PM, Jeen Broekstra <jeen.broekstra@gmail.com>wrote:

> A bit of general advice, by the way: regardless of which tool you choose,
> don't rely on published benchmarks too much to decide which tool "performs
> best". Try out several tools and see how well they work _for you_. Your
> requirements in terms of types of query, dataset size and form, etc. are
> likely different from any standard benchmark, and performance results may
> be significantly different.
>

It can be more useful to read the publications about the benchmarks than
the actual results :-)

Also, it is important to decide what the phrase "named graph" is going to
mean for this particular application.
RDF-1.1 basically says it's an IRI plus a graph; what that IRI denotes is
not specified;  if the application involves
"multiple named graphs" with "several relations among them", if the
relations are between the "named graphs", as opposed to between entities
mentioned in multiple graphs, there might be additional inferencing needed.

Simon
[RDF-1.1 changes statements from <Subject,Predicate,Object> to
<Subject,Predicate,Object,Ringo>]
Received on Thursday, 15 May 2014 23:36:12 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:41:42 UTC