Search Mailing List Archives
[protege-discussion] updating data model
jonathan.carter at e-asolutions.com
Fri Sep 19 02:28:42 PDT 2008
To some extent, it depends on how much change / refinement you expect
to do once you've branched off.
It it definitely worth having a look at the Prompt Tab in Protege,
focussing on the "diff" option. This provides a detailed report of the
difference between the frozen model and your branch. From the results
- which covers both Classes and Instances - you can choose which
differences to carry over and which to reject.
I think that this would work quite well as long as the frozen and
refined model are still strongly related. That is, your changes are
truly refinements and updates to the frozen model. Where I had trouble
was with the concept of merging in lots of changes from a branched
meta model (Classes) and a separate set of instances in the model - it
got too complex to manage - I had too many changes / possible clashes
in the Instances, basically (hundreds+ Instances) to bring across.
More a problem with the nature of my models, I think.
So, what I've been doing is to baseline my model (Class and Instances)
and then I've been using the Protege Script Tab to introduce
"packages" of changes to both the Classes and Instances, using Python
scripts. This is great for large numbers of changes and where I need
to be able to consistently (and sometimes 'intelligently') apply the
update to many instances of my model (I have several Protege models to
Having said all that, it sounds like - from what you've said - the
Prompt Tab approach would suit your needs best. Let your users
continue to use the data model that you currently have. Do a File-
>Save As on the Project and create your branch project. You continue
to work on the refined model and plan to have regular review of the
differences between the stable and the branch projects - taking
changes across to the stable instance as appropriate.
Hope this helps
Jonathan Carter - Head of Technical Architecture
Enterprise Architecture Solutions Ltd
Assess your EA maturity at:
On 19 Sep 2008, at 02:06, Jones, Josette F. wrote:
> We have followed the postings to the list with interest. Thanks
> they are mostly very instructive.
> My team and I are developing an ontology for the Geriatric Syndrome
> as part of a state funded grant for collaboration between the
> university and a local company. The goal of the ontology is to
> drive a semantic web application for automated textmining and DSS
> based on current research and local practices.
> We started in July and although progressing we are not finished
> yet. The company though wants to “freeze” in a datamodel. I’m a
> liitle wary of this since we have not extensively tested the
> ontology and its constraints.
> How can we continue to refine and test our model and update the
> datamodel? Any suggestions. Thanks for your time
> Have a gentle day!
> Josette Jones, RN, PhD
> Assistant Professor School of Informatics and School of Nursing
> Indiana University, Indianapolis
> 535 Michigan Avenue
> Room 491
> Indianapolis, IN 46202
> Phone: (317) 274-8059
> Fax: (317) 278-7669
> Email: jofjones at iupui.edu
> Webpage: http://informatics.iupui.edu/people/jofjones
> protege-discussion mailing list
> protege-discussion at lists.stanford.edu
> Instructions for unsubscribing: http://protege.stanford.edu/doc/faq.html#01a.03
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the protege-discussion