Stan, it does my heart good to see both the continuing relevance of that lengthy weekends discussions (and BBQ), and how much we actually got "right."
At 02:13 PM 10/28/2010, Blum, Stan wrote:
This "three-way" was the essence of the definition used in the ASC model.
<copied text, from http://wiki.tdwg.org/twiki/bin/viewfile/TAG/HistoricalDocuments?rev=1;filena me=Ascmodrpt.pdf page 24>
Entity Name: COLLECTING-EVENT (Supertype) Description: The act of collecting zero or more COLLECTING-UNITs at a particular LOCALITY and TIME.
</copied text>
Other CollectingEvent properties discussed included the conceptual equivalents of: CollectingMethod, Collectors (0-many), StatedDateTime, and StatedLocality (=verbatimLocality).
Note that the zero-or-more cardinality on collecting-units (covering Roger's observed absence) was discussed at some length and kept intentionally. (I'll try to find John Damuth's very funny proposal to establish the department of NULL collections at the Smithsonian.) I think observations were also discussed as a type of collecing-unit, but they weren't included in the draft and I don't remember why. Perhaps because the focus was on collections, and observations would have expanded the scope too much to be dealt with adequately. Also, the model did not include in a structured way was any measure of collecting (sampling) effort. That would have been relegated to text in a collecting method or collecting event remarks (inadequate for quantifying abundance).
Back to the issue of event definition: If Event is defined as just the conjunction (association, intersection, join) of space and time, there is nothing to tell you why this particular interval is of interest. From the old school information modeling perspective, the definition should say WHAT happened. In our biodiversity domain, it implies the act of trying to collect or observe and that implies a collector/observer and something collected/observed, including zeros. I see Joel just posted support for that notion.
And just to show that I'm not completely stuck in 1992, in the MVZ model -- a more detailed model for mammal, bird and herp collections, completed as recently as 1996 -- recognized a distinction between the number (count) of items observed and the number collected.
Our challenge is still to how to accumulate these artifacts of conceptualization in an organized way, and to record how much support there is for particular concepts.
Cheers,
-Stan
On 10/28/10 10:09 AM, "Arlin Stoltzfus" arlin@umd.edu wrote:
On Oct 28, 2010, at 11:58 AM, joel sachs wrote:
Hey Rich, Hilmar, Paul, and everyone -
I liked the definition from a couple of weeks ago:
"An occurrence is a tuple consiting of time, place, individual, and some optional properties."
What's that lacking?
Nothing, but the form of "occurrence" as a 3-way relation between a thing, a place and a time might be less amenable to formal reasoning than some other formulations, depending on how its rendered.
I thought that was the reason others had introduced "event" as place + time. Otherwise, what is the reason for "event"?
Arlin
Arlin Stoltzfus (arlin@umd.edu) Fellow, IBBR; Adj. Assoc. Prof., UMCP; Research Biologist, NIST IBBR, 9600 Gudelsky Drive, Rockville, MD tel: 240 314 6208; web: www.molevol.org
tdwg-content mailing list tdwg-content@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-content
tdwg-content mailing list tdwg-content@lists.tdwg.org http://lists.tdwg.org/mailman/listinfo/tdwg-content