[uf-new] Proposed Microformats: hRebuttal, hEvidence, hSource, hConclusion and hArgument

[Keith Alexander]

Danny Ayers wrote:
> On 24/05/07, Costello, Roger L. <costello at mitre.org> wrote:
>
>> I propose the creation of several Microformats for identifying the
>> parts of an argument.
>
> Coincidentally, Keith Alexander's been experimenting [1] in the same
> domain using an eRDF (i.e. upper-case Semantic HTML) representation of
> the IBIS (Issue-Based Information Systems) RDF vocabulary (one I put
> together a while ago, it certainly
> seems doable.
Actually, not a coincidence, but prompted by Roger's proposal, I just 
marked up a sample bit of html using Danny's vocabulary to see how it 
might be used together with the SIOC (semantically interlinked online 
communities) vocabulary.

I like the idea of the things that you could do with semantically 
annotated debates, but Colin has a good point that in most cases there 
is not a structure with any implicit semantics there to begin with. To 
implement this in a bulletin board system for instance, you would need 
to change, not only the templates, but the software behind the 
templates, in order to get users to enter in the extra information 
(whether they are rebutting, etc). Either that, or you require people to 
go to extra effort writing what is often spur-of-the-moment stuff.

  So you might well struggle to get people to publish  semantically 
structured arguments. That isn't to say it's a terrible idea though, and 
you may well find  niches in publishing behaviour where greater semantic 
structure has more immediate ROI than others. In fact, it's probably 
worth looking outside the realm of HTML as well, into other formats, 
such as DocBook, and TEI, where authorship is less impulsive, and 
semantic structure more valuable. (That said, the examples of e-mail 
mailing lists Danny gives in his vocabulary are also interesting).

If you want a quick way to develop and prove your idea, you could do 
worse than take a look at eRDF and/or RDFa, where you have the 
flexibility to describe anything you want, and you can don't need to 
write any new parsers (so your format can evolve as your ideas develop), 
and of course, Danny's IBIS vocab (and the existing work it is based on) 
are well worth a close look.

Then, if a more light-weight and domain-limited approach seems  
advantageous,  you can put up a  profile  uri for a custom format 
(perhaps incorporating a more tightly defined interpretation of the 
vote-link format), and get people to start using it. If you write a 
GRDDL profile transformation for it, it can be backwards compatible with 
any material you have  already published with eRDF or RDFa.

But if you wait for your format to be whole-heartedly embraced by the 
microformats community before you proceed, you probably won't get 
anything done at all.


Cheers,

Keith

-- 
Keith Alexander
http://semwebdev.keithalexander.co.uk/blog/