Unique Name Assumption

[Ed Shaya]

Alexandre RICHARD wrote:

> - if your field of application includes isotopes then you obviously 
> cannot use an individual for the carbon element

It is a very small subset of astronomy that worries about atoms but does 
not worry about isotopes.

> - if your field only requires the carbon element without anymore 
> details, then you can use an individual since if you create a class for 
> the carbon element it will have only one individual in its extension (in 
> fact in a case like this, from a strict DL representation you actually 
> should use an individual... but as usual the choice is often motivated 
> by performances factors such as the overall complexity of the DL used 
> and the kind of inferences to make)

I would think it is a little too early to know that there will only ever 
be one individual for each element.  Do we really know all of the future 
use cases?  Do you really want to cut off any possibility of another 
individual?
How about TheCarbonInMon1Cloud to refer to just the carbon atoms in a 
particular cloud?
What if you need to distinquish between CarbonInTheHaloOfM31 and 
CarbonInTheDiskOfM31?


> All of this to say that I understand your concerns, Ed, but if I am not 
> mistaken you are working on ontologies for the Astronomy field with an 
> objective of creating a rather complete and universal structure whereas 
> I am working on a very small, dedicated structure to be used as a 
> limited but fast and powerful tool for inference on astronomical object 
> types only. Certainly both of our works have numerous similarities, but 
> your limitations and issues are vastly different from mine, which is why 
> I sometimes have to disagree.
> 
It costs very little to have a Class Carbon and an individual named 
Carbon as well.  For this tiny cost you make your ontology more easily 
interoperable with laboratory scientists such as astrochemists, 
astrobiologists, physicists, and instrumentation developers.
If interoperability carries zero weight with your organization then I 
still would suggest that the elements be classes and perhaps the 
isotopes could be individuals (And be sure to make them allDifferent).
But, I think it is very odd to work in the field of semantics and not be 
somewhat interested in having interoperability with all of astronomy and 
physics at least.


> 
> Cheers,
> 
> Alexander
-------------- next part --------------
A non-text attachment was scrubbed...
Name: eshaya.vcf
Type: text/x-vcard
Size: 257 bytes
Desc: not available
URL: <http://www.ivoa.net/pipermail/semantics/attachments/20070419/6d3ad171/attachment-0001.vcf>