Re: LD4KD CFP - Linked Data for Knowledge Discovery Workshop at ECML/PKDD

On 2014-09-19 14:27, Mathieu d'Aquin wrote:
> Hi,
>
> [I responded on twitter in more or less the same line, so this is just a
> short summary]

Thanks for your response!

> I see that you are asking similar questions to many others, so I
> probably won't add much here. Basically, my answer is: I agree with the
> principles you are trying to put forward, but I don't think it is such
> an obvious thing to do. Workshops are for exchange between people. If
> the process of engaging with a workshop is made more complicated, it is
> kind of defeating the purpose. Also, this sort of things are happening,
> slowly, but it is a process which obviously would rather be bootstrapped
> in higher impact venues.

I tend to consider all "levels" of information exchange sufficiently 
valuable that they can make an impact what we are all working towards. 
But, I guess we can both agree that there is no need to debate on this 
particular matter.

> Finally, I see that you are part of the organising committee of
> SemStats, which as far as I can tell does also promotes PDF as the
> (only) publication format - maybe what you could do to help, in addition
> to sending emails to mailing lists, would be to show us the way?

SemStats, like some other workshops e.g., COLD, accepts (X)HTML+RDFa 
documents as long as they are consistent with their presentations in 
parallel to those in PDF. This is to encourage the use of LD/"Web 
friendly" technology stack (as opposed to the PDF, which is 
fundamentally intended for the desktop or print). Where and how the 
"publication" of the proceedings are conducted is entirely orthogonal to 
having source of the research documents as "machine-friendly".

While I am not personally in favour of seeing PDFs flying around about 
Linked Data, I do acknowledge the middle-ground given the state of the 
publishing workflow. I think the middle-ground for the time-being is 
that, research documents can be first published using the Web stack in 
an honest manner, and at the same time, create a PDF copy of that 
document (which is essentially created for "free") to fulfill the 
(hopefully) temporary requirements of the research venues.

See also: http://csarven.ca/call-for-linked-research

The question is, will the research venues, like the ones you are 
investing a lot time towards allow "Web friendly" submissions in 
addition to the PDF? Or will they continue to insist submissions only in 
PDF? What steps or responsibilities should the research venues about 
Linked Data take on in order to come slightly closer to the 
SemWeb/LinkedData vision?

Let me stress here again (as I do in almost every conversation) that, we 
are not dealing a technology problem here. It is absurd to even think 
that the Linked Data folks can not cope anything but LaTeX or their 
WYSIWYG editors. It is equally silly to think that there isn't 
sufficient or adequate tooling to essentially publish Web pages.

https://github.com/csarven/linked-research is only one modest way of 
dealing with these technological "problems".

There is valuable information in the public funded research on Linked 
Data that's being put forward, yet, we are defaulting to PDF, why? 
Because the publisher said so? The true publisher of the research 
documents are the authors, not the organizations that are laughing their 
way to the bank. ;)

> Thanks,
> Mathieu.
>
> ps: I usually don't argue on mailing lists, and will most of the time
> miss emails to a mailing list even if they are directly addressed to me,
> so consider this message an exception.

Acknowledged. My intention is not to "argue", but to have an open 
discussion. If this or similar mailing lists are not for that, I find it 
rather awkward to use this mailing list for the sole purpose of 
receiving CfP's in PDF about "Linked Data" ;)

Thanks again,

-Sarven
http://csarven.ca/#i