EML issues

[Matt Jones]

Hey Peter,

I'm starting to surface again after a bout of grant-writing. Sorry for 
the delay.

I think the working group for units is a great idea.  The current unit 
dictionary is good, but it lacks consistency and I believe it still has 
numerous errors.  Critically reviewing it, revising it, and adding as 
many new units as we can discover would be a great thing.

I can try to get someone from SEEK to follow through with the changes to 
the unit dictionary if the set of changes needed was clear.   It might 
take a bit to find enough time, but we'll get it done eventually.  It 
would be a good activity for the SEEK postdoc from the KR group who is 
developing ontologies, but since Rich Williams left that position is 
vacant. Maybe the position will be full by the Montreal meeting (if I 
get my act together and recruit!)  But if there were a volunteer from 
the LTER community who wanted to compile the changes into the STMML 
format of the unitDictionary and integrate it, that would be more than 
welcome.

We should (simultaneously) pursue the idea of a clearinghouse for unit 
definitions, but that can probably be folded into the idea of one or 
more ontology repositories as we have been discussing in SEEK.  The EML 
unitDictionary really expresses semantic relationships between units in 
quantitiative terms -- it represents a simple ontology unto itself, and 
it is nicely unambiguous because the relationships are quantitative. So, 
given that we want to be able to access remote ontologies from in the 
seek tools (like Kepler), we should be able to make sure this approach 
accomodates quantitative ontologies like the unit dictionary.

Thanks,
Matt

Peter McCartney wrote:
> Matt.
> 
>  Im in the IMEXEC meeting. yesterday we discussed the concern shared by
> several sites with expanding the units dictionary. I tried to summarize
> some of the discussion we have had about impoving how we handle
> enumerated lists and dictionaries in EML. there is an interest in
> holding a working group at the next IM meeting in montreal to compile a
> comprehensive list of unit definitions. Ive suggested that we follow
> Margaret OBrien's example of submitting a list in a bug which they agree
> is a reasonable simple approach, but would like to ensure that we can
> provide for someone to do the follow up to make sure they get into the
> dictionary. Is there someone specifically handling this that we can
> invite to the meeting, or should we be thinking of identifying someone
> to do that follow up? 
> everyone is interested in a future solution like DAT file refresh
> services, but recognizes that for now, doing a large batch update of the
> dictinary would benefit people. 

-- 
-------------------------------------------------------------------
Matt Jones                                     jones at nceas.ucsb.edu
http://www.nceas.ucsb.edu/
National Center for Ecological Analysis and Synthesis (NCEAS)
University of California Santa Barbara
Interested in ecological informatics? http://www.ecoinformatics.org
-------------------------------------------------------------------