Hi Alex,
Thanks for the discussion! I talked about fine-grained ontologies not because we're making one, but to get a better understanding of what a "middle-grained ontology" should be, and how we can provide a bridge between the coarse-grained document structure and the fine-grained structures.
ORB itself doesn't have a fine grained structure -- it's focused on the head, body, and tail, and refines the body with IMRAD:
http://www.w3.org/2001/sw/hcls/notes/orb/
Rather, there are a *lot* of existing fine-grained ontologies (such as the domain ontologies you mention, as well as document-focused and rhetoric-focused ontologies), and my goal is to figure out how to bridge between ORB's coarse structure and those fine-grained ontologies with a middle-grained ontology. Those "fine-grained ontologies" are not all the same kind of thing: they differ (ontologically!) because some are focused on the document, others on the process of science, and some on the reasoning ("why") behind the science.
At other times, the group has discussed *why* we need papers and not just better open data with the sort of e-science and workflow/provenance recording you discuss. Anita calls the paper "a story that persuades with data", and I think the main challenge is not about replacing the description of experiments (you're right--that's best served by domain-specific ontologies!). Rather, the challenge is in understanding and explaining how the experiment is *used* as part of the rhetoric: **why** should this particular data persuade a domain scientist?
I'm not exactly sure what you mean by "Once the structure of the document is solved". For instance, if we want to create a claims network (which many of us would find exciting and useful), we need to extract the claims out of the document. Yet those claims appear throughout the document -- they're not limited to the introduction, or discussion, or conclusion -- and can (Anita assures us) be repeated throughout the document. And it's not just claims but also evidence that will need to be extracted. I'd like to hear more about your thoughts on document structure -- because your understanding of what's needed could help focus efforts.
Thanks again, and let me know what you think!
-Jodi
On 11 Apr 2011, at 16:25, Alexander Garcia Castro wrote:
> Hi Jodi, nice talk.
>
> As a general question, shouldn't the description of experiments be something to be described by domain specific ontologies? Also, there are repositories for experiments, e.g Array express for microarrays; there are also minimal standards for reporting several experimental procedures, shouldn't the sections in the document make use of, pointing to, these resources? if ORB and other document related ontologies go into describing experimental information I am afraid it may be difficult to maintain and not very practical to implement.
>
> fine-graining ORB should, in my opinion, make use of other ontologies for addressing domain specific issues. for instance, if a researcher wants to go deeper into materials and methods most likely the information that will allow him/her to "replicate" the experiment will be living somewhere else in the web; also, most likely there will be an ontology for specific experimental procedures, measures, etc involved.
>
> Once the structure of the document is solved it is easy to use resources such as bio2rdf to resolve it against existing resources in the web. if not, then authors can easily point to these external resources -AO can easily deal with such a use case. my point is, experimental information is highly domain specific and there are now ontologies (OBI) and repositories for experimental information as well as efforts for standardizing minimal amounts of information (http://www.mibbi.org/index.php/Main_Page) for reporting specific experimental procedures, shouldn't the fine grained structure of ORB seek to make use of these resources?
>
> On Mon, Apr 11, 2011 at 8:09 AM, Jodi Schneider <jodi.schneider@deri.org> wrote:
> Looking forward to talking with you in about an hour.
>
> Here are slides for discussion -- about the overall goal of bridging between ORB and the fine-grained ontology with a middle-grained ontology.
>
> The other two PDFs show
> (1) the relationship between ORB and Anita's middle-grained ontology and
> (2) a first attempt to bridge between DExI (top-level) and Anita's middle-grained ontology.
>
> -Jodi
>
>
>
>
>
>
>
> On 10 Apr 2011, at 19:38, Waard, Anita de A (ELS-AMS) wrote:
>
> > Dear all,
> >
> > Monday's scientific discourse call will focus on an update of the medium-grained ontology alignment by Jodi Schneider: see agenda and call details below, or at http://www.w3.org/wiki/HCLSIG/SWANSIOC/Actions/RhetoricalStructure/meetings/20110411
> >
> > Kind regards,
> >
> > Anita
> > Anita de Waard
> > Disruptive Technologies Director, Elsevier Labs
> > http://elsatglabs.com/labs/anita/
> > a.dewaard@elsevier.com
> >
> > Meeting April 11, 2011, 10 am EST, 3 pm BST, 4 pm CET
> > [edit] Agenda:
> >
> > * Jodi Schneider: 'Medium-grained ontology and alignment'
> > * Status of ontology development
> > * AOB
> >
> > [edit] Dial-in & IRC Information
> >
> > * Dial-In #: +1.617.761.6200 (Cambridge, MA)
> > * Dial-In #: +33.4.26.46.79.03 (Paris, France)
> > * Dial-In #: +44.203.318.0479 (London, UK)
> > * Participant Access Code: 42572 ("HCLS2")
> > * IRC Channel: irc.w3.org port 6665 channel #HCLS2 use IRC direct link or (see W3C IRC page for details, or see Web IRC)
> > * Mibbit quick start: Click on mibbit for instant IRC access
> > * Duration: 1hr
> > Elsevier B.V. Registered Office: Radarweg 29, 1043 NX Amsterdam, The Netherlands, Registration No. 33156677 (The Netherlands)
> >
>
>
>
>
>
> --
> Alexander Garcia
> http://www.alexandergarcia.name/
> http://www.usefilm.com/photographer/75943.html
> http://www.linkedin.com/in/alexgarciac
> Postal address:
> Alexander Garcia, Tel.: +49 421 218 64211
> Universität Bremen
> Enrique-Schmidt-Str. 5
> D-28359 Bremen