From: Pankaj J. <pj...@co...> - 2009-01-20 18:42:23
|
Barry Smith wrote: > At 08:16 AM 1/20/2009, Norman Morrison wrote: >> Hi Chris / OBI Group, >> >> In sending this post, you've actually provided a use case that >> potentially opens up a (small) can of worms that I've also been thinking >> about for a while. >> >> When including an has_habitat <-> is_habitat_of relation, I think it's >> worth stepping back for a moment and considering whether or not an >> has_environment <-> is_environment_of relation might be more appropriate? >> >> The reason being is that the definition of habitat amongst >> biologists/ecologists - more often than not - implies the USUAL or >> PREFERRED environment of an organism. ie If you were visiting Milwaukee >> County Zoo, you wouldn't expect to see the habitat of 'Zero' the polar >> bear described as 'a compound in a zoo' on his plaque, rather you would >> expect to see something like 'Arctic ice sheets'. As you point out this >> might make more sense for habitat to be considered as type level relation. >> >> I was actually considering a proposal to substitute all instances of the >> term 'habitat' in EnvO with 'environment', in order to avoid any further >> confusion. > > Note that we need instances, too -- otherwise > EnvO won't work with GAZ. That is, EnvO should > supply the resources to state that Zero is > currently living in the XYZ compound in the ABC > zoo. I think that where Zero is living is then an > instance of the type environment (an organism is > living in it), even we are not talking here about > the usual environment for arctic bears > BS > > As I see it, I think this relationship is driven by the annotation/metadata rather than hard coding it. As Norman points in his latest mail, existence of the individuals representing a species found in a wild can be hard coded, but not for the individuals introduced by the human subject/as the climate changes, habitats move and sometimes new inhabitants adapt or move as well. We can say that X-species is prevalent (is this a good word?) in a given GAZ location in association to a given set of Envo terms. However, for a given individual (accession/stock/germplasm/seed) it is all annotation driven. Because although it was first identified/collected from the original source, the later generations may have been maintained/bred at a different location. Pankaj |