Re: Biological Taxonomy Vocabulary 0.1

----- "Toby A Inkster" <tai@g5n.co.uk> wrote:

> From: "Toby A Inkster" <tai@g5n.co.uk>
> To: "Kjetil Kjernsmo" <Kjetil.Kjernsmo@computas.com>, "Semantic Web" <semantic-web@w3.org>
> Sent: Wednesday, October 8, 2008 9:10:01 AM (GMT+1000)
> Subject: Re: Biological Taxonomy Vocabulary 0.1
>
> On 7 Oct 2008, at 10:03, Kjetil Kjernsmo wrote:
> 
> > I'm a latecomer to this discussion, and not a biologist, so sorry  
> > if this is
> > silly, but I thought Uniprot had a pretty uptodate taxonomy of  
> > species?
> 
> 
> The Biological Taxonomy Vocabulary takes a completely different  
> approach from UniProt, uBio, etc.
> 
> Look at FOAF. The FOAF spec doesn't contain instance data - that is, 
> the spec doesn't define URIs for all the billions of people alive on 
> Earth. Instead, it defines a few dozen terms like foaf:name, which  
> other people can use to create instance data. Similarly, Dublin Core 
> doesn't define URIs for millions of books, but does define a small  
> number of properties which are useful for describing books (and  
> films, music, etc).
> 
> Existing biological ontologies have tended to focus on defining URIs 
> for tens of thousands of different species. The Biological Taxonomy  
> Vocabulary doesn't do that - instead it defines a smaller set of  
> terms (less than twenty) that people can use to define species  
> themselves.

Is it reasonable to recast the current taxonomy ontologies in terms of this vocabulary or is it designed to prevent that happening in some way?

What is stopping someone using the Uniprot defined taxonomy predicates to define their own instances? What do you gain in particular by not using the existing Uniprot predicates which are currently already used to describe a large number of known species? There is no restriction on utilising their predicates in any particular other ontology AFAIK.

Cheers,

Peter

Received on Tuesday, 7 October 2008 22:28:35 UTC