[Taxacom] Finding your species in Mendeley

Paul van Rijckevorsel dipteryx at freeler.nl
Thu Dec 8 02:26:06 CST 2011

From: "Jim Croft" <jim.croft at gmail.com>
Sent: Wednesday, December 07, 2011 10:00 PM

+1 on the 'manually'. It will take years, in fact we have already
spent years on it, but it is a finite and doable task.

The key issue is one of trust. We might use an item or link a computer
found but we will check it. Entries by competent humans, on the other
hand, are 'trusted' and considered verified, i.e. the link has been
followed and the right thing is at the end of it. If we are going the
do the 'let the computer find an add stuff' thing, a 'verified flag is
critically important.

Nomenclators are the foundation of everythng we do and as such they
are not allowed to be wrong. They can be incomplete as a work in
progress, even ambiguous if the ambiguity is explained, but they are
not allowed to be wrong.

Find a mistake in APNI, if we can't convince ourselves it is not a
mistake, a pleasant and sentient human WILL fix it... Find something
missing and we will unmiss it, but this might take a little longer...

This sounds great as a statement of principle, but still, over time 
I have come to associate the APNI component of IPNI as the
least useful of the three, with the least depth and the most errors.
Also it is a pity that there does not seem to be a way to have
errors in APNI corrected, as opposed to the IK (easiest) or
the GCI.


More information about the Taxacom mailing list