Search Mailing List Archives


Limit search to: Subject & Body Subject Author
Sort by: Reverse Sort
Limit to: All This Week Last Week This Month Last Month
Select Date Range     through    

[protege-discussion] Protege 3.3.1 DB Schema

Jonathan Carter jonathan.carter at e-asolutions.com
Thu Feb 19 06:54:51 PST 2009



   Hi Dennis,

   Makes sense. I don't have any documentation for the structure that  
is used, although there are only about 7 or 8 columns in the table and  
you are right you can decipher what each does from the source code. (I  
had an extensive look around in there trying to resolve the SQLServer  
2005 issue.

   Have a look at the code in the protege/storage/database classes:

    
http://smi-protege.stanford.edu/svn/protege-core/trunk/src/edu/stanford/smi/protege/storage/database/

   In particular, the class of interest used to be  
DatabaseFrameDB.java but things have been refactored. To see what's in  
Protege 3.3.1, go back to a version in the subversion repository  
that's about January 2008:

    
http://smi-protege.stanford.edu/svn/protege-core/branches/namespace/src/edu/stanford/smi/protege/storage/database/DatabaseFrameDb.java?rev=8708&view=markup

   Hope this helps

Regards

   Jonathan

   Quoting "O'neill, Dennis" <doneill at marathonoil.com>:

> Jonathan,
>
> We have been using the XML export to handle our ontologies so far.
> However, the issue we are dealing with is that the one of the ontologies
> I currently have is too large for MS SharePoint to process efficiently
> as an XML import. For certain uses (SharePoint libraries), we only need
> a subset, and hence want to filter the entries before XML export/import.
>
> Having looked at the DB that is generated, I can see how it is really an
> export of the Frame meta-database; but that is exactly the thing we want
> to decipher. If I understood the format, perhaps I could post-process it
> into a true ontology DB (tables for class instances, for example) that
> we could then do our queries over.
>
> Is there any documentation along these lines, or could you point me at
> the code that does the DB creation?
>
> Thanks,
>
> Dennis
>
> Dennis M. O'Neill
> SAIC
> Contractor to Marathon Oil Company
> Enterprise Content Management
> Room 2068D
> (281) 236-5690 (Cell)
> doneill at marathonoil.com
> -----Original Message-----
> From: protege-discussion-bounces at mailman.stanford.edu
> [mailto:protege-discussion-bounces at mailman.stanford.edu] On Behalf Of
> Jonathan Carter
> Sent: Thursday, February 19, 2009 2:47 AM
> To: User support for Core Protege and the Protege-Frames editor
> Subject: Re: [protege-discussion] Protege 3.3.1 DB Schema
>
> Hi Dennis,
>
> From my experience of the database schema, I would advise against
> operating directly (even reading) the underlying database table (note:
> singular).
>
> Protege database is basically a meta-meta-database. So, all your classes
> and instances and everything are just instances of the meta-meta
> classes, e.g. Frame.
> (This is roughly speaking - I haven't spent much time on the details of
> this). It works really nicely but is rather complex to interpret (to say
> the least) with a lot of references to other rows to follow to get the
> picture that you have in Protege (the GUI editor).
>
> When Protege sets up the database it puts some indexes in to help with
> all the look ups required. (Unfortunately, there's an issue with MS
> SQLServer 2005 which means that Protege 3.3.1 can't create all the
> indexes)
>
> What I would recommend - and what I use for analysing the ontology with
> an external application - is an XML rendering of the ontology, using the
> format that the experimental XML format from the File->SaveAs menu.
> I've written a tab widget to snap-shot the knowledge base in this XML
> format and send it to my analysis application.
>
> The great thing about the XML format is that although all the entries in
> it are pretty much Classes and Instances (in fact, I just capture the
> Instances) it's very much easier to interpret - and this could be used
> to feed a database that is more directly representative of your
> application / ontology if XML processing is not an option for your
> user's application.
>
> Regards
>
> Jonathan
>
>
> Sent from my BlackBerry(r) wireless device
>
> -----Original Message-----
> From: "O'neill, Dennis" <doneill at marathonoil.com>
>
> Date: Wed, 18 Feb 2009 17:02:41
> To: User support for Core Protege and the Protege-Frames
> editor<protege-discussion at mailman.stanford.edu>
> Subject: [protege-discussion] Protege 3.3.1 DB Schema
>
>
> _______________________________________________
> protege-discussion mailing list
> protege-discussion at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/protege-discussion[1]
>
> Instructions for unsubscribing:
> http://protege.stanford.edu/doc/faq.html#01a.03[2]
>
> _______________________________________________
> protege-discussion mailing list
> protege-discussion at lists.stanford.edu
> https://mailman.stanford.edu/mailman/listinfo/protege-discussion[3]
>
> Instructions for unsubscribing:
> http://protege.stanford.edu/doc/faq.html#01a.03[4]
>

  _______________________________________

Jonathan Carter  - Head of Technical Architecture
Enterprise Architecture Solutions Ltd

Web: www.enterprise-architecture.com
_______________________________________

Assess your EA maturity at:
www.enterprise-architecture.com/EAvaluator
_______________________________________

W O R L D    C L A S S    A R C H I T E C T U R E



Links:
------
[1] https://mailman.stanford.edu/mailman/listinfo/protege-discussion
[2] http://protege.stanford.edu/doc/faq.html#01a.03
[3] https://mailman.stanford.edu/mailman/listinfo/protege-discussion
[4] http://protege.stanford.edu/doc/faq.html#01a.03

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.stanford.edu/pipermail/protege-discussion/attachments/20090219/0b897ffe/attachment.html>


More information about the protege-discussion mailing list