[uf-discuss] species microformats & OpenSearch

[Shorthouse, David]

> .could ever contribute to the semantic web in a meaningful way & will
stand
> the test of taxonomic revisions
I agree with this. It's unclear to me how the current proposal even
relates to the research gathered, and what use cases it might support.
Typically, microformat proposals are heavily influenced by the
analysis of examples collected.  I've tried doing this work at
<http://microformats.org/wiki/species-examples-regrouped>.

Most of the useful examples look similar to one of the sites you mentioned:
<a
  href="/data/spiders/14441"
  onMouseOver="window.status='';return true"
  title='Click for species description'>
     <i>Aculepeira carbonarioides</i>
     (Keyserling, 1892)
</a>

Looks to me like most mentions of species don't contain much
information about them, but rather link to to another page that does.
To me this resembles tagging, where species mentioned is the tag, and
the endpoint of the url is the resource representative of the tag.
[David Shorthouse wrote:] 

Indeed, this is a lot like tagging and are nothing more than links to other
species pages in an attempt to permit users a chance to quickly get to
information about other species in the same Genus as the one on the
currently visible page. Also in a trivial way, this is to permit search
engine "spiders" a chance to navigate the thousands of pages.

What is actually more useful from a taxonomic standpoint are the "Synonyms
and Other Recognized Nomenclature" tables on each species page that are the
1:1 mappings of historic nomenclature to that currently recognized. These of
course also have LSIDs.

Perhaps with further analysis, we can modify hReview or xFolk to be
useful for species, in order to model what is actually happening in
the market.

Can you:
* elaborate on the kinds of use cases you would expect a species
microformat to support
[David Shorthouse wrote:] 
What I would ultimately hope for is a means to aggregate such species pages
across multiple resources. This means some sort of scaffolding that is
intelligent enough to know that a species with a microformat for Lycosa
fuscula Thorell, 1875 and another for Pardosa fuscula (Thorell, 1875) refer
to the very same species. A browser plug-in that found species microformats
on the page could highlight these & provide something like a floating div to
indicate current nomenclature in the event that a tagged name is not the
currently recognized name. This would permit one unfamiliar (or familiar)
with the species an opportunity to quickly recognize that the provided page
may have useful information, but that the nomenclature is dated.

* confirm whether or not the above model is the most common way of
publishing species mentions
[David Shorthouse wrote:] 
In fact, there is no model. The vast majority of similar species pages have
no common ground, no tagging, and are merely free-form text with images.

* collect intances of the authoritative resources and their markup of
the species
   * what is the most commonly published information (on the authoritative
end)
[David Shorthouse wrote:] 
These would be peer-reviewed publications, most of which are paper-based.
The ICZN and other organization have their rule-set for what constitutes a
new species name, but there is little in place to programmatically tap into
that data, though many organizations are making steps toward that ultimate
goal. Speaking about spiders, the authoritative work for their nomenclature
is the World Spider Catalog:
http://research.amnh.org/entomology/spiders/catalog/INTRO1.html, which is
essentially an HTML representation of a paper-based publication with no
means to programmatically tap into the data.

   * how is it represented (on the authoritative end)
[David Shorthouse wrote:] 
Unfortunately, I can't speak to that.

Ben