W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > July 2006

RE: LSIDs and ontology segmentation

From: Xiaoshu Wang <wangxiao@musc.edu>
Date: Fri, 14 Jul 2006 09:49:23 -0400
To: "'Chimezie Ogbuji'" <ogbujic@bio.ri.ccf.org>, "'w3c semweb hcls'" <public-semweb-lifesci@w3.org>
Message-ID: <000601c6a74c$50d0eba0$22241780@bioxiao>

> I'm glad you brought that up.  I do think we have the 
> expressiveness in OWL (and in Description Logics,
> generally) to model term alignment (equivalent classes/properties,
>   inverse roles, value restrictions, etc..) as a seperate 
> ontology This would refer to (and therefore owl:import) the 
> HCLS-Experiment ontology ( to ground the mapping semantically 
> not just for annotation purposes ).

Yes, you can use DL, and if necessary to invent new logical terms, to merge
ontologies.  But the problem is where you would place these matching
statements?  In BOSS, in Gel, or in HCLS-Experiment?  If you placed in any
of those, you essentially bind them together, either old or new.  Then after
a few turn overs, we would be left with a huge bloated ontology, containing
a lot of statements that most won't care.  The issue is an analog to a
typical use case in software development of hard-coded dependency.  Depedent
codes are difficult to maintain and almost impossible to evolve in the long
run.  Ontology engineering is not much an exception here.  But developing
software, sometimes we can start with approach to let the dirty code running
first.  But I don't think it is a good idea in ontology engineering.  We
have to think about what it is going to be 5 or 10 years down the road.

Xiaoshu
Received on Friday, 14 July 2006 14:24:21 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 18:00:44 GMT