Re: Fwd: Using Statement Identifiers to Manage Provenance

Paul Groth wrote:
 > Some interesting background info around implementing provenance in
 > triple stores.

This seems like rehashing an old topic to me.  I'm not seeing anything here that 
isn't covered by reification.  (The complaint about reification triple-bloat is 
really a complaint about one particular implementation technique.  I think most 
serious triple stores that actually handle reification use an internal quad 
mechanism.)  The real problems with reification are (a) the lack of a clear 
semantics, leading to early implementations with divergent semantics, and (b) 
the main use-cases for reification seem to be more readily addressed by named 
graphs.

My current sense is that RDF community consensus favours named graphs:
(1) the SPARQL syntax makes explicit provision for querying named graphs,
(2) the current RDF working group is giving consideration to including a 
mechanism to encode named graphs within a single RDF documemt
(3) even when using RDF without named graph support, named graphs map directly 
to a natural web-based implementation: RDF documents retrievable from web URIs.
(4) many RDF triple-store implementations include named graph support at the API 
level, which may also be accessible via SPRQL queries (I know Jena/TDB supports 
this)

My sense is that named graphs are a more convenient way to apply provenance 
information.  I would suggest that, in typical use, it is not single RDF triple 
to which one wants to attach provenance information, but a collection of triples 
from a common source.

#g
--

Paul Groth wrote:
> Some interesting background info around implementing provenance in 
> triple stores.
> 
> -------- Original Message --------
> Subject: Using Statement Identifiers to Manage Provenance
> Date: Tue, 17 May 2011 16:51:54 +0200
> From: Makx Dekkers <mail@MAKXDEKKERS.COM>
> Reply-To: Makx Dekkers <mail@MAKXDEKKERS.COM>
> To: DC-PROVENANCE@JISCMAIL.AC.UK
> 
> Using Statement Identifiers to Manage Provenance
> http://www.bigdata.com/bigdata/blog/?p=254
>

Received on Wednesday, 18 May 2011 10:59:40 UTC