Re: RDF's challenge

Hi all,

I'm sure that on this particular list, everyone will agree that RDF is
web-native (of course it is). As a member of OKFN myself I have found
Rufus' quote very hard to defend (certainly not as a LOD researcher).
Nevertheless it was very thought provoking and sadly it may be a very close
representation of the reality we are currently in.

In our lab (MultiMedia Lab at the university of Ghent) we did a short
brainstorm over lunch. These were the outcomes (rather problem statements):

 * People who should re-use LOD are data amateurs. They are user-experience
experts and great front-end engineers, but they do not know what the
difference between URL, URI or IRI is, they have no clue what HTTP stands
for and REST is something they do at night.
 * People think RDF is an immensely complicated technology and overthink it
all the time.
 * There are no killer use cases / demos of LOD (apart from
http://everythingisconnected.be, but that is only using 1 dataset, it
doesn't even follow external links) to convince policy makers to start
publishing LOD. Most of the sparql endpoints in the LOD cloud are academic
use cases that proved a point and then stopped working.
 * When you ask windows users to install a program on a debian machine,
they will start a browser and start looking for a binary on the web while
everyone perfectly knows that apt-get is the way to go, right? Similarly
when you ask a 12 year old to write a program which processes all the
toilets in a certain city, (s)he will start looking for a CSV file on a
search machine instead of going to the cities' linked data portal and fire
a SPARQL query. How can we start making this mind shift?

Personally, I believe LOD Portals will play a key role in the future
evolution towards Linked Open Data from both sides (it will be the subject
of many of my next papers). What Rufus says certainly holds true today: RDF
is not the first thing organisations think of when publishing data today,
and LOD is not the first thing developers think of for their projects
today. I'm sure however that time will bring more use cases, more tools,
and more great show cases of LOD so that it will become more of a reflex
rather than a step in the dark.

Kind regards,

Pieter




On Tue, Jun 11, 2013 at 7:36 PM, Kingsley Idehen <kidehen@openlinksw.com>wrote:

> On 6/11/13 12:58 PM, Hugh Glaser wrote:
>
>> Nicely put, David.
>> I have heard people going the other way and disconnecting them, however.
>> That is, suggesting that Linked Data does not need to be RDF, which I do
>> find confuses people (and me!)
>>
>
> It isn't all or nothing.
>
> It just means you don't always have to start with RDF when trying to
> explain or demonstrate the virtues of Linked Data.
>
> RDF's real contribution lies in enabling machines to understand the
> relationship semantics that are used to construct the web-like structured
> data.
>
> You can construct 5-Star Linked Data endowed with coarse-grained (rather
> than fine-grained) entity relationship semantics without any knowledge of
> RDF whatsoever. All you need is TimBLs original meme (which outlines how to
> use HTTP URIs to enhance structured data representation) or fundamental
> knowledge of how to use pointers to enhance structured data representation.
> BTW --  many Web developers actually have this kind of knowledge without
> every digesting an RDF related spec.
>
> Again: I make these comments to encourage flexibility in our approaches to
> Linked Data evangelism etc... Linked Data is a powerful conduit to RDF and
> beyond. It doesn't have to be confined to RDF -- since they are both useful
> in their own rights.
>
> There's also lots of RDF out there where none of the IRIs resolve. They
> still deliver value, even if said value doesn't scale to the World Wide Web.
>
> Kingsley
>
>
>> On 11 Jun 2013, at 16:56, David Booth <david@dbooth.org>
>>   wrote:
>>
>>  On 06/11/2013 10:59 AM, Kingsley Idehen wrote:
>>>
>>>> [ . . . ]  many RDF advocates
>>>> want to conflate Linked Data and RDF. This is technically wrong, and
>>>> marketing wise -- an utter disaster.
>>>>
>>> I have not heard RDF advocates conflating Linked Data and RDF, but maybe
>>> you talk to different RDF advocates than me.
>>>
>>> AFAICT, the vast majority of RDF advocates know that Linked Data is RDF
>>> in which URIs are deferenceable to more RDF, but RDF is not necessarily
>>> Linked Data, because RDF itself does not require URIs to be dereferenceable.
>>>
>>> David
>>>
>>>
>>
>>
>>
>
> --
>
> Regards,
>
> Kingsley Idehen
> Founder & CEO
> OpenLink Software
> Company Web: http://www.openlinksw.com
> Personal Weblog: http://www.openlinksw.com/**blog/~kidehen<http://www.openlinksw.com/blog/~kidehen>
> Twitter/Identi.ca handle: @kidehen
> Google+ Profile: https://plus.google.com/**112399767740508618350/about<https://plus.google.com/112399767740508618350/about>
> LinkedIn Profile: http://www.linkedin.com/in/**kidehen<http://www.linkedin.com/in/kidehen>
>
>
>
>
>
>