Re: [Graphs] Proposal for Named Graph Semantics

Antoine Zimmermann wrote:
> I propose something that goes very much in the same direction as Alex's 
> proposal:
> 
> Definition(Graph map) A graph map GM is a partial function from the set 
> of IRIs to the set of g-snaps.
> 
> Definition(Temporal graph map) A temporal graph map TGM is a partial 
> function from [-inf,+inf] which maps a time point to a graph map.

Nice, but if we define a mapping from a time point to a graph map, then 
how do you handle the case where a time point maps to multiple graph 
maps (for instance when you have auth* to consider)?

> There exists a special temporal graph map, called the HTTPmap, such that 
> at any given point in time t, HTTPmap(t) maps a URI to the parsed RDF 
> graph of the document retrieved via an HTTP GET of the URI at the time 
> t. (if HTTP GET does not provide an RDF serialisation, then the mapping 
> is not defined on that URI).

As above, doesn't consider auth*. At a single instant several different 
graphs may be associated with a specific URI.

Thoughts?

Best,

Nathan

Received on Friday, 8 April 2011 22:28:26 UTC