RE: Antwort: RE: Semantic web article in Nature Biotechnology

- Robert

> One thing we have to remember is that biologists are building 
> ontologies to do a job of work. they are not produced as some 
> end of CS or SW research. the annotation of biological data 
> with terms from the gene ontology, for instance, has done 
> much for computational analysis of data -- and it is not 
> using sW technology.
>
> We can't just expect people, whose priorities are not those 
> of the SW, to simply use this technology because we say it is 
> good. It is up to us to demonstrate its utility. Such 
> demonstration should be done in collaboration with biologists 
> -- only in this way can we ensure sensbile biology is done 
> and SW technologies are used appropriately. RDF and OWL are 
> not easy technologies and we ask a great deal of people when 
> we expect strict esemantics to be used in a complex domain.

Yes.  I agree completely.  There is actually a gap between the view of
ontology for CS people and for biological people.  As you said, biologists
are building ontologies to do a job but not to answer a philosophical or
logical questions.  The ontology in bioligist's eyes are more of a treaty
than logical representation, that in CS view is on the reverse of that view.
It needs dialog to bring the view to a middle ground and mechanisms to
stretch to both directions. 

> Again, as Phil mentions, the OBO format is essentially 
> compatible  with OWL semantics and  converters are available. 
> there is, for instance, a Protege plugin and we have one 
> under development here in Manchester. One of the rules of OBO 
> is that their ontologies should be orthogonal. If they are 
> not, then perhaps this is a sign that such a goal is 
> difficult and that we should use our tools and techniques to 
> resolve the problem. Finally, if we want nice modular, SW 
> technology ontologies then perhaps we should produce usable 
> tools and techniques for delivering them. Modularisationof 
> OWL ontologies is still a hot topic. I note in your latest 
> email "because I am still chewing on some problems." -- so it 
> is not a surprise bio people are not doing it as a priority 
> for their work?   

Well, I am not blaming them. That is why I am sending request to this list
to put up good best practice guideline. 

> I think the point of imports and modularisattion etc. are  
> mtherhood and apple pie, but until the technologies and 
> methodologies (how do I fragment my ontology appropriatly for 
> any possible re-use?) are in place and yield  tangible 
> benefits for ontology developers then it will not be done. 
> What bene fit do I, as an ontology developer, get from 
> spending a lot of precious development time making my 
> ontology re-usable. Obbiously  I know the answers, but we 
> need to make it cheap for people to do such nice engineering 
> solutions.

Yes.  We don't differ much in our opinion. :-)  But IMHO, how to partition
ontology will always arbitrary. But that shouldn't stop people from doing
it.  The benefits of partition ontology actually benefit for the ontology
developers.  It helps their ontology be shared across and it also help them
not to borrow others.

Received on Monday, 3 October 2005 15:46:31 UTC