Annotating our specs with RDF...

Okay.... I was talking with Manu yesterday and I decided we should 
really be annotating our specs with RDF - via RDFa of course.  I then 
decided I should just extend ReSpec.js (the framework we are using to 
publish) to just do this for us magically.  So far, so good.  I went in, 
made changes, got it all working, and hit a wall.

This is embarrassing, but after all this time I STILL don't really grok 
RDF.  Here's an example.  Our documents have a title.  dcterms has a 
title term in its vocabulary.  Great!  So, I want to express in RDF that 
this document has a title of whatever our title is....

<title about="" property="dcterms:title">our title</title>

That *seems* right to me.  I will get a triple that has a subject of the 
current document, a predicate of title, and an object of the title 
string.  Is that right?  Is that how a real RDF semweb person would do this?

What about the Abstract?  The bibo vocabulary has an abstract term.  
Great!  Our sections are surrounded by 'div's, so....

<div id="abstract" resource="#abstract" rel="bibo:abstract"><h2>....

Is that right?  Am I saying this document (the about is inherited) has 
an abstract at some URI?  That seems backwards to me.  It seems more 
sensible to say:

<div id="abstract" resource="#abstract" rev="bibo:abstract"><h2>....

This URI is the abstract for this document...  Or.... maybe the thing 
shouldn't be a pointer at all?  Maybe it should be the actual abstract?

<div id="abstract" property="bibo:abstract"><h2>....

I struggle with similar things when trying to annotate Chapters (bibo 
has BookChapter, for example), references (dcterms has 'reference' but 
do we annotate the link to the reference, or do we annotate the actual 
reference in the References section?  And what is a subject and what is 
an object?  Help?

-- 
Shane P. McCarron                          Phone: +1 763 786-8160 x120
Managing Director                            Fax: +1 763 786-8180
ApTest Minnesota                            Inet: shane@aptest.com

Received on Sunday, 4 July 2010 17:55:48 UTC