Re: Toward easier RDF: a proposal

On 11/21/2018 7:38 PM, Holger Knublauch wrote:
> This sounds like a potentially very worthwhile discussion.
> 
> On 22/11/2018 8:40 AM, David Booth wrote:
>> One bright light in our favor is that RDF already provides a
>> very solid foundation to build upon, based on formal logic.
> 
> Would you mind clarifying this statement a bit? What practical benefits 
> would the foundation on formal logic add to a future (simplified) RDF, 
> for average users? I have seen plenty of evidence that some aspects of 
> the semantic technology stack are being regarded as too academic, and 
> that the role of formal logic has been one driver of this detachment. 
> Related topics are the non-unique-name-assumption and the open world 
> assumption that are typically neither understood nor expected by average 
> users.

Let's face it: formal (predicate) logic as usually used and presented is 
difficult and repellent for most normal people.  Whether this is because 
of the terminology, the formality, or whatever, it's very hard.  OTOH, 
many engineers learned to use boolean algebra for the design of logic 
circuits, and to write SQL statements some minimal understanding of 
logic and sets is needed.  So there's some hope.

With RDF, there is also the added difficulty of understanding what an 
"interpretation" means, as a practical matter.

I would say that removing or lowering this barrier would be a big help. 
An RDF stack, and conventions and terminology for practical use that hid 
these aspects (or didn't require a user to consider them) would go a 
long way to making the whole thing more palatable.

Received on Thursday, 22 November 2018 12:25:27 UTC