Re: Advancing translational research with the Semantic Web

Matthias wrote:
> > The specific difficulties with the location of objects and  
> > processes are to some degree inherent in the current OWL versions  
> > of BFO and the Relation Ontology. They make it a bit difficult to  
> > create statements that relate processes to certain locations and  
> > force you to make statements about the participants of the process  
> > instead.

Alan wrote:
> Could you elaborate on this? My understanding is that according to  
> BFO, pretty much everything about a process is determined by the  
> participants (the process is existentially dependent on the  
> participants). 

Yes. But there are some situations where we might prefer to talk about the 'location' of the process directly, without explicitly talking about its participants.

Let me introduce the relation <is_location_of_process> as an example.
<A> <is_location_of_process> <B> means that <A> contains all of the participants that make up process <B>.
A concrete example for the use of this property would be

<Human_organism> <is_location_of_process> <blood_circulation> .

The property would have two main advantages:
1) It does not force us to make statements about the participants of a process when we are unsure what the actual participants are. In the example above, we would probably be forced to introduce a fiat part of the human organism like 'human_organism_circulatory_system' that is not very clearly defined. We had a similar problem in our ontology of neuronal morphology with the localization of the process of ionic currents.
2) It gives us the possibility to say that the physical extent of a process is CONFINED to the extent of a certain object. Normally, the process of blood_circulation is confined to the spatial extent of the human_organism, unless you are undergoing blood dialysis. This property would allow us to state this fact a bit more easily in OWL with its open world assumption.

This problem is probably not significant enough to warrant the introduction of such a new relation, but at least it is a little difficulty that many people working with the Relation Ontology will encounter. Maybe it should be addressed in the documentation of the ontology.

cheers,
Matthias Samwald

----------

Yale Center for Medical Informatics, New Haven /
Section on Medical Expert and Knowledge-Based Systems, Vienna /
http://neuroscientific.net


.
-- 
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail

Received on Wednesday, 30 May 2007 14:44:51 UTC