Re: is linked data about RDF or EAV or just structured data?

On 8/8/12 2:50 AM, Wilde, Erik wrote:
> hello kingsley.
>
> On 2012-08-08 3:10 , "Kingsley Idehen" <kidehen@openlinksw.com> wrote:
>> A tree is a rooted graph. At the base you'll find a graph.
> so you're following henry's reasoning that because everything somehow can
> be mapped to a graph, everything naturally "is" a graph? again, i'd advise
> you to look at the IETF registry and take a look of what's there and see
> how much sense some of these formats would make when represented as a
> graph in a generalized graph notation. more importantly, anything can be
> mapped to anything, but this doesn't tell us anything about the things
> we're mapping.

I agree with Henry, for sure.
I don't really agree with your characterization of the points we are 
making here re. Linked Data.

We are saying the representation of a Data Object is a graph pictorial 
in 3-tuple form.

Data is how Subject Observation is Represented.

When you speak of structured data you are ultimately speaking about 
semantics and syntax for expressing subject observation.

Linked Data is about webby structured data. The whole thing is grounded 
in existing architecture of the world wide web (awww).

Linked Data is about a webby approach to the following re., structured data:

1. entities or things are described by web documents
2. each entity has a name, ditto its attributes and attribute values
3. each document has an name/address (an address serves as a web 
document name)
4. each document is comprised of content
5. the content takes the form of an eav/spo triples based graph
6. entity names resolve to description document content.

The baseline model for structured data representation is an 
entity-attribute-value or subject-predicate-object triple. This is old. 
What's new is the combined effects of combining URI abstraction 
ingenuity and structured data representation.

>
>>> - there is the "Linked Data is based on RDF" perspective which is shared
>>> by most people, then
>> It isn't but I can't force you to accept this view point. I am confident
>> that in due course you will realize why this is the case.
> simply google for "define linked data" and see what people are writing.

I am not disputing the existence of mass confusion. I am desperately 
trying to kill the propagation of mass confusion.
>
>>> - there's the "linked data is just data that's linked on the web"
>> Yes, and then you have to hone into the definition of data and how its
>> represented.
> that's where you lose me. all we need for this generalized view is a way
> how data can link to other data, so any data that has the capability to
> include links qualifies. all REST needs is the capability for clients to
> find hypermedia controls in representations.

A client can figure out everything about a chunk of data (a data object 
resource) via the relationship semantics that constitute its graph 
pictorial based representation. Basically, curl -I -X OPTIONS is all an 
HTTP based user agent should require to start discovering what's 
possible via operations and URIs.

Example:

curl -I -X OPTIONS http://kingsley.idehen.net/dataspace/person/kidehen
HTTP/1.1 200 OK
Server: Virtuoso/06.04.3132 (Linux) x86_64-generic-linux-glibc25-64 VDB
Connection: Keep-Alive
Allow: 
GET,HEAD,POST,PUT,DELETE,OPTIONS,PROPFIND,PROPPATCH,COPY,MOVE,LOCK,UNLOCK,TRACE
Date: Wed, 08 Aug 2012 13:11:45 GMT
Accept-Ranges: bytes
Content-Type: text/html
DAV: 1,2,<http://www.openlinksw.com/virtuoso/webdav/1.0>
MS-Author-Via: DAV
Content-Length: 0


Ultimately, WebID ACLs which determine how identities are mapped to 
specific operation privileges etc..

As I keep on saying, if we just loosely couple Linked Data and RDF, you 
will find that self describing documents of immense sophistication are 
already here. RESTful interaction is also here, but in more fine-grained 
form since a human doesn't need to read a mime type document for figure 
out what capabilities are accessible to a user agent, instead, this 
would simply be a case of understanding a data 
dictionary/schema/ontology/vocabulary  that defines the classes, 
properties, individuals in the entity relationships graph for their 
particular discourse domain.


>
>> Linked Data is about whole data representation via an
>> Entity-Attribute-Value data model enhanced with de-referencable URIs.
>> Linked Data uses de-referencable URIs to denote observation subjects,
>> their attributes, and attribute values (optionally). It also uses the
>> same URIs to identify web resources and then, via indirection (explicit
>> or implicit),  it associates a URI with a web resource that bears the
>> description of said URIs referent.
> you keep repeating this without ever substantiating it.

Its already substantiated in existing literature. And to save time, read 
TimBL's design issues note, but just drop the references to RDF and 
SPARQL from #3, which is also the state of the original version of the meme.

> i have never seen
> this definition of linked data before, and just repeating that this is
> what linked data really is does not help to make the case why you think of
> it that way.

Well, you aren't searching hard enough, unfortunately.

>
> and like i said earlier, we as a group should not spend too many cycles
> (re)defining linked data, this is not what we have been chartered to do.

But that's exactly what you are trying to do by seeking what you present 
as a loosening of the definition of Linked Data en route to making it 
RESTful.

You are not really getting my point. That's why we are looping. You are 
not picking up the semantics of Linked Data and its specific nuances 
because you see it as inextricably bound to your understanding of RDF, 
which in of itself is actually inaccurate.

I am burning this much time on this matter because of I've done this 
dance a few times, and at some point, the nuances I am trying explain 
will be revealed.

BTW -- I've been dealing with data issues (client-server patterns, 
middleware and databases) for 20+ years. REST is just a contemporary 
tweak of client-server computing patterns. The moniker itself is an 
attempt to explain the ingenuity of URI abstraction applied to 
client-server computing, exemplified by the HTTP protocol.

Decouple RDF and Linked Data and you will see all the glories of REST 
and beyond, courtesy of the ingenuity of URIs and the HTTP protocol.


To conclude, the goal of Linked Data is to solve the fundamental 
challenges associated heterogeneous and disparate data access, at 
web-scale, via Web oriented entity relationships.

This goal of the World Wide Web has always been about using hyperlinks 
(URIs) to create and disseminate: data, information, and knowledge.

If you tweak your perception of RDF and Linked Data you will find what 
you seek :-)

>
> cheers,
>
> dret.
>
>
>


-- 

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