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

wangxiao wrote:
> Thank you for the input.  I know of (but not KNOW in the matter of
> detail) about the ontologies you mentioned.  But a few reasons that
> "push" me off about these ontologies for the use in the context of
> Semantic Web are: (1) They do not comply to SW technology
> specifications. 

Most of the ontologies do "comply" with biology however. Content
is much more important that formalism, to my mind. The formalism
can always be fixed.


> If you go to OBO and check those developed ontology, most of them are
> not coded in RDF/OWL.  

Most of the OBO ontologies are in "OBO format". The translation to
OWL is a syntactic one in most cases. 

> Even for a few that have the "OWL-version",
> they are not coded corrected.  For instance, GO has a presumed
> namespace "http://www.geneontology.org/owl/", but access that URI
> leads to a 404. 

The namespace in GO is a URI not a URL. Should there be any 
expectation that it does anything other than 404? 


> (2) Many of them are overlapping, for instance, GALEN and MSH, etc.
> IMHO, ontology merging is no easier than schema integration.  It is a
> transformation of the old problem but not the solution.

This is, of course, true. 


> (3) All these ontologies are still developed in a monolithic manner.
> No consideration about the granuality and practicality has been
> emphasized.  

There have been many considerations of practicality made. Monolithic
development is a problem, but both OBO and the MGED ontologies have 
been attemping to ensure consistency between ontologies as well as 
minimise overlap. 

> An encyclopedic-like ontology is only useful as referece
> but not for other "reasoning task".  I have used Jena before to play
> around the w3c's example Wine Ontology.  Reasoning with subClass
> etc., is O.K. but once all other axioms are considered, my computer
> always run of memory. 

We routinely reason over ontologies much larger than the wine ontology.
Semantic web technology much scale to the size of the "encyclopedic" 
ontologies; otherwise, what use is it? 
 
> If an ontology is designed
> in such that it can not be executed in the context of SW, what good
> it does to us - at least in this group? 

I think that these ontologies can be used in a semantic web context.
For example, see our own gohse, which makes use of GO. 

http://cohse.man.ac.uk/gohse/

> My own personal interest in
> ontologies are about experimental data, like how to model gel, MS
> data etc.  Maybe, as a group, we can start from this, I already have
> some prelimiary work done like a trivial top ontology about
> scientific research - BOSS
> (http://www.charlestoncore.org/ontology/boss), which I don't think
> overlap with any of the ontologies you recommend? 


The boss ontology looks, at a quick glance, fairly similar to the 
parts of the provenance ontology from mygrid, or perhaps the 
experiment ontology by Larisa Soldatova and Ross King, or even the 
work on Hybrow. MGED also has an experimental ontology somewhere
within it, or, at least, I think that this is the case. 

Cheers

Phil

Received on Tuesday, 27 September 2005 09:46:21 UTC