[seek-kr-sms] Taxon/KR integration prototype proposal

[Robert A. Morris]

Minor point from a lurker:

In the prototype proposal, you write:

To answer these questions, the researcher must gather and use several 
pieces of information, including:

•	information about taxa (e.g., what are the non-human primates?);
•	information about geography (e.g., where is South America?);
•	information about environmental properties (e.g., climate);
•	information about ecological interactions (e.g., what types of 
parasitic relationships exist between ants);
•	and real data (e.g., climate data sets and presence/absence data sets).


The third bullet (information about environmental properties) is not a 
pre-requisite to answering the sample questions, but rather an artifact 
of the particular predictive tool (GARP) that gets offered by the 
hypothetical system. Later, in the step-by-step the issue rearises where 
it needn't. Namely in step 4, all except 4c ("select [stuff related to 
GARP]") could meaningfully be accomplished before a choice of analytical 
engine.

In other words, I think it is possible to distinguish choices to be made 
before the modeling tools are chosen from those arising as a consequence 
  of that choice. Thus in the introduction quoted above, "[choose] 
information about environmental properties" represents something to be 
done /after/ the modeling regime is selected, whereas the other bullets 
represent issues that could (must?) be resolved /before/ it is selected. 
 From an engineering point of view, I guess that distinction will need 
to be reflected in the architecture.

I'm not close enough other than to speculate that it's possible SEEKers 
are overfocused on GARP, and that GARP's way of doing business sometimes 
intrudes on the generalities. (Oops. Self-invited visitor waggles finger 
at admired hosts. Sorry.)


thau at learningsite.com wrote:

> Hello everyone,
> 
> Shawn and I have been working on describing a prototype which would
> integrate the things the ks-sms folks are doing and the things the taxon
> group folk are doing.
> 
> Here's what we have so far.  Please take a look at this, and answer a few
> questions for us:
> 
> 1.  Would a prototype like this be useful?
> 2.  Is it scoped about right to have something to show by early May?
> 3.  Are there other, better use cases integrating taxon and sms stuff?
> 4.  The milestones are rough - but any thoughts about important
> 	things that should go in there would be appreciated.
> 5.  Does anything in this not make sense?
> 6.  What kind of beer?
> 
> Enjoy!
> Dave and Shawn

-- 
Robert A. Morris, Professor of Computer Science
University of Massachusetts at Boston
100 Morrissey Blvd; Boston, MA 02125
http://www.cs.umb.edu/~ram http://www.cs.umb.edu/efg
phone: (+1)617-287-6466 fax:   (+1)617-287-6433