Re: Linked Data discussions require better communication

On 6/21/13 8:02 PM, David Booth wrote:
> On 06/21/2013 10:25 AM, Kingsley Idehen wrote:
>> On 6/21/13 10:15 AM, David Booth wrote:
> [ . . . ]
>>> The only sensible interpretation of the stars is that they indicate
>>> milestones of progress *toward* "Linked Open Data" -- *not* that there
>>> are five levels of Linked Open Data.
>>
>> That makes sense. Thus, why can't you accept the same thinking if we
>> look at RDF unique selling points as part of such a journey too?
>
> Because just as the goals of the web cannot be achieved by having "a 
> journey toward URIs", the goals of the Semantic Web cannot be achieved 
> by having "a journey toward RDF". 

Look, the journey has zilch to do with labels. It has everything to do 
the the concepts they denote. The destination is an Internet-scale mesh 
of web-like structured data with varying degrees entity relationship 
semantic fidelity. That's the journey, the literals "RDF" and "Semantic 
Web" are must identifiers for which specs handle the name-address 
indirection in our human minds.

It's the concept that matters.

> RDF is *fundamental* to the Semantic Web, just as URIs are 
> *fundamental* to the Web.  RDF is the universal data model that 
> enables Semantic Web data to be meaningfully combined by automated 
> applications.  That *cannot* be done without either: (a) a boat load 
> of artificial intelligence and processing power that is out of reach 
> of most mortals; (b) a dramatic new discovery that the world has not 
> yet seen; or (c) a universal data model.
>
>>
>> What's wrong with folks arriving at points in the continuum where RDF's
>> virtues kick-in without actually being aware of RDF?
>
> Nobody has claimed that people must be *aware* of RDF for a document 
> to be standards-interpretable as RDF.  Indeed, it seems very likely 
> that *many* JSON-LD users will be unaware that JSON-LD is actually RDF 
> in addition to being JSON.  The important point is just that the data 
> *be* standards-interpretable as RDF.  Whether or not it *looks* to the 
> untrained eye like RDF is quite irrelevant.

Again, I defer with you when you say "standards-interpretable as RDF" 
since (to me) that statement is quite ambiguous. What does "as RDF" 
mean? Does that characterization trump the fundamental concept of an 
Internet- and Web-scale mesh of entity relationships where:

1. each entity is denoted unambiguously using a resolvable identifier 
(or reference e.g., an HTTP URI);
2. each entity is associated with another via a relationship ;
3. each relationship is represented via a 3-tuple (triple) statement;
4. each member of the relationship has a specific role (subject, 
predicate, object OR  entity, attribute, value OR object, sign, 
interpretant);
5. each relationship is a member of set known as a mathematical relation;
6. each mathematical relation (in this context) has a predicate that 
determines membership.

>
>>
>> BTW -- I still don't know if you accept the world view outlined in my
>> venn diagram [1]. I don't want to misquote you, so at the very least,
>> could you confirm if you agree with the venn diagram or not.
>
> No, I do not.

Thank you for making that clear. Our differences are now much more 
simpler to understand.

Related:

1. https://en.wikipedia.org/wiki/Sign_relation - Sign Relation (some 
background on this subject matter).


Kingsley
>
> David
>
>>
>> Links:
>>
>> 1. http://bit.ly/16EVFVG -- how Structured Data (Linked Data), Predicate
>> Logic (RDF), and Identifiers (URIs) are related.
>>
>
>
>


-- 

Regards,

Kingsley Idehen 
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen