Re: Datatypes (Was: Re: Consumer guidance)

Ivan,

On 23 Nov 2011, at 16:15, Ivan Herman wrote:
> On Nov 23, 2011, at 15:53 , Jeni Tennison wrote:
>> On the other hand, the argument is that useful consumers always have built-in knowledge about the vocabulary that they understand, so they know what datatypes to expect for each property. Given that, relying on publishers to supply a datatype for each value is problematic because (a) they might get it wrong, by assigning an incorrect datatype or no datatype at all, so a consumer always has to fix up those mistakes anyway and (b) it gives publishers more work to do when we want to make their lives easy.
> 
> Yeah... but that means you have to have vocabulary aware processors all the way down. If I push the data down a datatype aware inference engine, for example, (say, Pellet) then somebody has to fill in the missing bits. I understand that it some cases that is possible, of course. And it may require a not-always-obvious interpretation of the data

I don't think that it means you have to have vocabulary-aware processors all the way down. But you do need them at some point for most useful applications. So as a vocabulary-aware consumer you might have something like:

  1. parse into RDF (no vocabulary knowledge needed)
  2. insert into triplestore (no vocabulary knowledge needed)
  3. perform SPARQL query to extract data (need vocabulary knowledge)
  4. display results as HTML (need vocabulary knowledge)

As the consumer, you could use your knowledge of the vocabulary to do some conversion when you do the SPARQL query, or before you put the data into the triplestore. There's no need for the publisher to provide you with datatyping information.

Cheers,

Jeni
-- 
Jeni Tennison
http://www.jenitennison.com

Received on Wednesday, 23 November 2011 18:29:48 UTC