RE: ISSUE-66: LinkedDataT

On 6 Aug 2014 at 18:09, Ruben Verborgh wrote:
>>   This combination of structure and URLs is the essence of Linked Data
>> What exactly do you mean by *structure* in this context? Could we
simplify
>> it by saying "This usage of URLs is the essence..."? Or do you think we
>> would lose something if we did?
> 
> I tried to refer back to the sentence above:
> 
>     Machines prefer structured data using unambiguous identifiers.
> That closes the story about Linked Data then.
> Perhaps this can be done clearer?
> 
> Note that the structure is indeed important:
> HTML pages have a lot of URLs too, but we can't do the same things.

What about simply repeating it as in

  Such structured data with unambiguous identifiers is the essence...


>> I think you've done a good job of phrasing this.
> 
> Thanks :-)

Indeed, really great work!


>> Note, however, that there is a working group (CSVW) with a charter a 
>> "CSV file with thousands of numbers" useful as Linked Data.
> 
> Oh great, now this thread will explode again ;-)
> 
> On a serious note: any other example there would also work;
> I just want to indicate the difference between machine-processable and
> -interpretable.

I think this is fine as-is as you mention the missing context explicitly:

   ... Most of them are written in natural language, which machines
   cannot understand yet. And even if a document contains machine-
   readable information, **the appropriate context is often missing**.
   For instance, what do thousands of numbers in a comma-separated
   file mean?

AFAICT, the CSV WG is addressing this by adding such a context. Gregg, out
of curiosity, is the current idea still to leverage JSON-LD contexts to do
so?


--
Markus Lanthaler
@markuslanthaler

Received on Wednesday, 6 August 2014 21:33:57 UTC