[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: plant DB - prose, searches




> From: John Schinnerer <eco_living@yahoo.com>
>
> This is good news about the AAAS DB, at least where non-living (e.g.
> recorded in some static form) information is concerned.  For an in-use
> legal model of "open" documentation, see:
> 
> http://opencontent.org/
> 
> And...I'll go on with related stuff from several rich recent posts...
> 
> Given the fragility of the technology pyramid we reside at the top of
> and the huge differences between "living" knowledge and "dead"
> information, I often think that much of the effort and infrastructure
> that goes into high-tech projects such as we're proposing (and PFAF,
> AAAS, etc. are undertaking) would be better used to revitalize living
> knowledge.

Too right. My spin on the subject, is that I'd like to gather as much
of this living knowledge as possible and aid the communication of this
knowledge. What I'd like to do is create a forum where people can 
exchange this knowledge. A collaborative plant database could do this
very well. One very important thing is making it collaborative, trying
to get through the not-invented-here pheromone. Providing appropriate
credit for info added to a database is very important. But a sense of 
ownership is also vital to make the project work.

The AAAS project does have a very specific aim which is to stop people
locking up knowledge/information. We've seen lots of problems with 
patenting of plants which can go prevent people even using their
living knowledge.

> > The problem with writing things as prose is that they are difficult
> > to search on.
> 
> However, if the DB is full-text searchable, it can be done.  Main
> trouble is as with text-based web searches (which are often a full-text
> search on the search engine's entire indexed web space) - lots of
> spurious results that aren't relevant and have to be manually culled.
> Advanced search features (booleans, etc.) help but can still be
> frustrating and require more user expertise.

Free text searches leave a lot to be desired. The words used to describe
something might not match the words used to search for something. There
can be a lot of irrelevant information returned by many of the search 
engines. Context is important, acid in terms of soil might not be what
your looking for if you are looking for acidic berries. One of the
weakest parts of the pfaf db is the range info, which is a text fields.

Latin Name		Common Name     Range
Abelmoschus esculentus  Okra		The original habitat is obscure.; 
Abelmoschus manihot     Aibika		E. Asia - Southeastern Asia to
                                        Northern Australia.; 
Abelmoschus moschatus   Musk mallow	S.E. Asia.; 
Abies alba              Silver fir	C. and S. Europe; 
Abies amabilis          Red fir		Northwestern N. America - Alaska
                                        to Oregon.; 
Abies balsamea          Balsam fir	Northeastern N. America -
                                        Newfoundland to Virginia, west
to Alberta,
                                        Michegan and Ohio.; 

Looking for plants which grow in a specific state or country from this
information is practically imposible.

> I am in process of redoing the static resource links on my site as a
> searchable object DB (using Zope's ZCatalog objects and custom
> classes), including canned queries and keyword and full-text searches.
> This should give some indication of how object based DBs handle search
> issues.

	ttfn

		Rich
-- 
Plants for a Future: 7000 useful plants
Web:       http://www.pfaf.org/  or  http://www.comp.leeds.ac.uk/pfaf/
Main Site: Blagdon Cross, Ashwater, Beaworthy, Devon, EX21 5DF, England
           Tel: (+44 845) 458 4719
Email:     webmaster@pfaf.org (web related queries only)
Why not join Friends of PFAF and help our project grow
see http://www.comp.leeds.ac.uk/pfaf/friends.html for details.
PFAF electronic mailing list http://groups.yahoo.com/group/pfaf