[seek-dev] RE: resultset question

[Peter McCartney]

Well if you visit http://seinet.asu.edu you can search eml files for
dataset, literature, and protocol . We hope to have models up soon,
although there is no official model schema in eml 2 yet. 

If you select some spatial datasets (search on land use or landuse to
find some vector data, search on ndvi or savi to find some raster data)
and select to open them in the spatial visualization, then you will see
a very simple example of how the geometry and measurement scale
information in eml is used to guide a map renderer to dynamically build
a color renderer for the variable you chose. There are still some kinds
to work out - I know it will work well if you try it with some vector
data, we've had some problems still rendering the raster data but these
are ArcIMS problems we think. 

This map vis tool gives a good perspective on where our articulation
with semantic mediation might occur. In our current app, it shows you
the fields listed in the eml and asks you to choose which one you want
to see displayed on the map. The next step requiring semantic mediation
woujld be for you to be able to say - "I want to see a map of landuse
classification" and have the app be able to recognize from the semantic
tags that the variable you want to see rendered is the one labled "z93"
or whatever.