Re: Trust, Context, Justification and Quintuples

Andrew Newman wrote:
>>Using "s,p,o" as unique identifier is out of
>>the bounds of the RDF datamodel.
>>
> 
> 
> I'm not sure you should be too hamstrung by what is and isn't in the RDF
> data model when creating an efficient way of storing/modelling RDF.  The
> constraints of RDF can be applied on top of a much more efficient model.
> This means most of the time, I think, creating a simpler model not a more
> complicated one like creating a surrogate key from the s,p,o or making URI
> Refs, bnodes and literals all the same type and making triples/statements
> out of that.
> 
> 

That's fine for a particular implementation to to internally.  The real 
trick is how to get the corresponding information into an interchange 
syntax so someone else can make use of it.

Cheers,

Tom P

Received on Saturday, 20 December 2003 13:10:30 UTC