[tcs-lc] TCS/LC Name Domain

[Robert K. Peet]

Hi Paul,

1. To identify is to link an observation to a concept.

2. The observation can be vouchered or not vouchered.

3. The list of birds I have observed around my home is a list of
identifications. A determination label on a museum specimen is an
identification, even though the specific concept being linked to might be
unclear. A publication that lists the taxa known from a specific
geographic area is a list of identifications if and only if there is a
statement as to which field guide or other authoritative set of concepts
the names are being linked to; otherwise it is a list of concepts (albeit
rather vague, as in the ITIS list).

4. An identification can have attributes, such as fit.

5. An identification event can link more than one concept to an
observation, if two or more concepts were deemed by the identifier to
match, even if at different levels of fit (this is a more common situation
with continuous phenomena such as hybrids or ecological communities than
with regular taxa). I find it valuable to allow linkage to concepts that
are identified as not fitting too, so as to more precisely document the
identification event.

6. For qualifying components of an identification event, I have found it
useful to use the five fit values employed by the fuzzy logic community
[see S. Gopal and C. Woodcock. 1994, Theory and methods for accuracy
assessment of thematic maps using fuzzy sets. Photogrammetric Engineering
and Remote Sensing 60(2):181-188]: Absolutely wrong, understandable but
wrong, reasonable or acceptable answer, good answer, absolutely correct.



Bob Peet



---------------------


On Mon, 18 Apr 2005, Paul Kirk wrote:

> please define an identification ...
>
>
 ======================================================================
     Robert K. Peet, Professor & Chair         Phone:  919-962-6942
     Curriculum in Ecology, CB#3275            Fax:    919-962-6930
     University of North Carolina              Cell:   919-368-4971
     Chapel Hill, NC  27599-3275  USA          Email:  peet at unc.edu

                   http://www.unc.edu/depts/ecology/
                 http://www.bio.unc.edu/faculty/peet/
 ======================================================================