[open-science] (Lack of) data sharing for fossil data
pm286 at cam.ac.uk
Fri Apr 15 15:11:34 UTC 2011
On Fri, Apr 15, 2011 at 3:44 PM, Lance McKee <lmckee at opengeospatial.org>wrote:
> Thank you, Peter. That helps me put what I know a little about in
> perspective with respect to what I don't know anything about.
This century is all about blundering into areas we know nothing about, and
making connections. That what the OKFis very good for.
> I would only argue with you that "where" matters in the study of fossils,
> trees and many other objects of study not generally thought of as subjects
> of the geosciences. Studying such objects very often requires managing
> geospatial information, and so students of those Earth features would
> benefit from leveraging the work that I described below.
> Completely agreed. We are developing an application to extract data by
text-mining atmospheric publications. It involves
* chemistry (we have a tool OSCAR which is very good at that). It turns the
textual chemistry into CML molecules.
* geo-coordinates. Here we can certainly use help
* authors, titles, so we use OKF Open Bibliographic approaches (mainly RDF)
* proper names (survey stations, campaigns, places, etc)
* physical science (quantities and units - "a velocity of 20 km.hr",
"concentration of 12pptV" "temperature of 25 Celsius"). We use CML for this
For each facet we have to use different tools.
BTW if you have applications where you wish multidiscplinary data extracted
from text then OSCAR is able to do quite a bit (
http://blogs.ch.cam.ac.uk/pmr/2011/04/15/oscar4-launch-roundup/ ). There's
no magic, just hard work fitting the bits together.
> Curiously, lung science (
> http://www.technologyreview.com/computing/37387/?nlid=4352&a=f) and brain
> science (
> require, or will require, standard ways of dealing with spatial data that
> are not too different from the OGC's Earth-referencing geospatial standards.
> It's common that an abstraction in one domain can transfer to another. But
not universal - no golden hammer.
And what's the hardest thing?
Being allowed to extract our (i.e. the scientific community) own data from
our own publications without being sued. It's a sad fact that one of the
miain functions of the OKF is to enlighten the world that almost everything
is not open and to encourage them to work towards it.
Not enough people care yet.
Reader in Molecular Informatics
Unilever Centre, Dep. Of Chemistry
University of Cambridge
CB2 1EW, UK
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the open-science