[NHCOLL-L:487] Re: natural history collections database

[Emery, Rob]

(First up, apologies to Doug for what was intended to be a private email
finding its way to NHCOLL)

Doug is correct in saying that our output fields a pretty limited.  We
wanted it this way so as not to scare off amateur and student entomologists
or interested members of the public for that matter.  This is why we have
random images and so on.  Also, we don't believe it is good to arrive at a
web page and being confronted with a sea of choices, instructions,
disclaimers, directions, alternatives.  We try to keep the interface as
clean as possible and this is why we keep the users guide elsewhere for
those brave enough to look closer.

Our database was started about six years ago and it was only after we had
about 50,000 specimens databased that we thought it might be a useful
resource if put on the Internet.  We find the Internet interface handy for
quickly locating a specimens especially as this can be done on any computer
anywhere but if we need more detail we go to the master Access database.  I
suppose we could add a "more detail" link to each specimen record on our
webpage which could then go back to the database and return full details for
that individual specimen.  This is what we do in our Insect Pest
Identification database at
http://www.agric.wa.gov.au:7000/ento/pestweb/default.idc

As an aside, our web page filters out records of quarantine significance.
For example an insect not occurring in Australia might be detected on a ship
by one of our quarantine inspectors.  The ID would be confirmed by our
quarantine entomologist and the specimen might find its way into our
collection.  The last thing we would want is for one of our trading partners
to search our database for this quarantine pest and find a record that
indicates it occurs in an Australian port!

Doug asks below how we deal with placenames...
http://www.agric.wa.gov.au:7000/ento/_fpclass/icdb28.gif shows the screen
that automatically looks up lat/longs.  Int this example the specimen label
has "3 km SW of Weebubbie" but no lat/long.  If the lat/long was on the
label we would enter this information exactly and note that the information
was derived from the collector.  If there is no lat/long, we can enter
"Weebubbie" as the nearest named place and see what the lat/long database
returns.  In the above example "Weebubbie road blowhole" and "Weebubbie
cave" are the choices. In some cases the name occurs in two Australian
states so we have to pick the right one.  It comes back with a fuzzy match
so you can pick from a list of close matches.  If you misspelt "Weebubbie"
it would probably come back with nothing, so you would suspect something was
wrong and try again.  When you pick your nearest place off the list, the
form will replace what you typed with the correct name, this way we ensure
that the nearest named place is always spelt correctly regardless of the
literal label info which goes in the Locality field.  The database will
record that this lat/long came from another database and was not taken from
the label.  We also use a navigational aslgorithm which corrects for
lat/long offset (E.G. 3 km SW) if one is given.  Our offset can only be
calculated from a distance and direction, Biolink and EGAZ (I think) can do
fancy things like calculate the offset based on distance between two places
and distance along a particular road.

This has been an interesting, largely unfunded project for us, we have over
95,000 specimens databased to date which is nearly half our pinned
specimens.  We'll think about our slide and spirit collections later on.
The database is a tool for our agricultural entomology work but we like to
that that we can give something back to the entomological community by
making it available over the Internet.  The nice part is that new
information available on our website every day without the need for us to
edit any webpages.


--------------------------------------------------------------
Rob Emery
(Entomologist)

ph:     61 8 93683247
Fax:    61 8 93683223
Mob:   0419 935 991
mailto:remery at agric.wa.gov.au
http://www.agric.wa.gov.au/ento

Agriculture Western Australia
3 Baron-Hay Court,
South Perth, 6151
Western Australia





-----Original Message-----
From: dyanega at pop.ucr.edu [mailto:dyanega at pop.ucr.edu]
Sent: Thursday, 9 March 2000 4:58 AM
To: NHCOLL-L at lists.yale.edu
Subject: [NHCOLL-L:486] Re: natural history collections database


>*All* information on the label is stored in the database.  Too bad you
>didn't take Doug's bet regarding lat/longs because you would have a little
>more spending money.

Actually, I was - as you yourself noted below - going on only what one sees
when one accesses your website and queries the database. The output fields
are pretty limited, and there is no immediate indication there to tell the
user that there is much more information in the database than is visible
from the query interface. I'll note that I'd also intended this as a
private message, largely because I *wasn't* sure I was seeing all there was
to it, and I was hoping to find out a few more details privately.

>We also
>record the nearest named place, automatically taken from a gazetteer of
>Australian placenames.

Automatically? But if a name is misspelled, or if there are multiple places
with the same name, certainly no automated process is going to be capable
of resolving the problem?

>The worrying thing for me about Doug's comments is that he thinks our
>database is a spreadsheet-style database.  Doug seems to think that because
>we choose to present the information over the Internet in a user-friendly
>spreadsheet-type format that this is how the information is stored in our
>database.

Like I said, when you're in the query interface you're not told what sort
of structure and content exists in the database behind it all. Unless you
dig around elsewhere on the website, you're left guessing as to whether
that's all there is, and how it's entered and organized. I apologize if it
came across as an accusatory tone, rather than inquisitive - I probably
could have answered some of those questions myself if I'd poked around more
extensively.

Peace,


Doug Yanega        Dept. of Entomology         Entomology Research Museum
Univ. of California - Riverside, Riverside, CA 92521
phone: (909) 787-4315 (standard disclaimer: opinions are mine, not UCR's)
           http://insects.ucr.edu/staff/yanega.html
  "There are some enterprises in which a careful disorderliness
        is the true method" - Herman Melville, Moby Dick, Chap. 82