Search Mailing List Archives
Natasha F. Noy
noy at stanford.edu
Mon Aug 23 14:37:08 PDT 2010
> I'm curious about the distinction between views and extensions/
> imports. Our structure is such that:
> - 'oboe' imports 'oboe-core', extending the core ontology with
> additional classes.
> - 'oboe-sbc' imports 'oboe' (as well as 'oboe-units' and 'oboe-
> context') and defines subclasses of various classes defined in
> What is the recommended approach for registering these ontologies?
> Some options, as far as I can tell, might be:
> 1. Register each individual owl file as a distinct ontology
> 2. Register a top-level ontology (oboe-core) and then create views
> off of it (oboe, oboe-sbc**)
> **how is the "grandchild" view treated (ie, oboe-core -> oboe ->
> 3. Register a somewhat-top-level ontology (oboe) and create views
> off of it (oboe-sbc)
It's a very good question. I would suggest that it would make more
sense to register these as separate ontologies rather than the views
on a core ontology -- although any of the two approaches is valid. We
*can* handle grandchildren views, at least in our infrastructure, but
the UI may be less kind to them at the moment.
Your question brings an interesting issue though and maybe we should
consider "groups" of ontologies to facilitate these types of scenarios
in the future. Thank you for the idea!
More information about the bioontology-support