ISO 11179 and ontology modeling

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

ISO 11179 and ontology modeling

Thomas-60

Dear Protege User(s),

I'm currently working on a project which purpose is to
implement an ISO 11179 compliant registry of standard
data elements (in a pharmaceutical clinical
environment).

Based on what I know from the ISO 11179 standard,
registered data elements could be classified in an
ontological manner...

Is anybody aware of a succesful implementation of such
a standard, including ontological aspects ?

Thanks in advance for your help.

Kind Regards,

Thomas

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Crowley, Rebecca
Thomas:
 
I think I can interpret your question in two ways - and I am not sure which is the right way.
 
The Cancer Bioinformatics Grid (caBIG) a project of the NCI has a ISO/IEC 111179 compliant registry that links Data Element Concepts, Valid Values, Representation Terms, and other administered components to concepts in a DL based ontology - NCI Thesaurus. NCI has made some extensions to the ISO/IEC standard - which I think they have proposed back to the standards organization for linking to an ontology in this fashion -this is not currently required in ISO/IEC 11179 compliant registries.
 
You can read more at:
http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
Most everything you would need that they use is open source except for the current vocabulary server.
 
Several of us (inside and outside NCICB) who worked on implementing a data standard (the CAP protocols)  using caDSR and NCI Thesaurus are just finishing a paper that describes the whole semantic linking thing. Once we are done with it - I'd be happy to forward it.
 
But... if what you are proposing is to actually implement the ISO standard as a frame system in protege - I think that would be very useful...I sort of half-heartedly attempted this myself last year but ran out of gas while reading through the standard :-) If that's what you mean -- Cecil Lynch might have done something like this...Cecil are you out there?
 
R Crowley
University of Pittsburgh

________________________________

From: [hidden email] on behalf of Thomas
Sent: Thu 2/9/2006 3:46 PM
To: [hidden email]
Subject: [protege-discussion] ISO 11179 and ontology modeling




Dear Protege User(s),

I'm currently working on a project which purpose is to
implement an ISO 11179 compliant registry of standard
data elements (in a pharmaceutical clinical
environment).

Based on what I know from the ISO 11179 standard,
registered data elements could be classified in an
ontological manner...

Is anybody aware of a succesful implementation of such
a standard, including ontological aspects ?

Thanks in advance for your help.

Kind Regards,

Thomas

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html




-- Attached file removed by Ecartis and put at URL below --
-- Type: application/ms-tnef
-- Size: 5k (5702 bytes)
-- URL : http://protege.stanford.edu/mail_archive/attachments/16-winmail.dat


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

dcarlson

Hi Rebecca,

I am also interested in taking a deeper look at your iso 11179 work.  My focus
is on the e-commerce domain, where several industry groups are using 11179
(parts 2 & 5, related to naming and classification) to standardize the naming of
reusable "Core Components" for orders, invoices, etc.

I took a quick look at your site, but have you created an OWL ontology for
representing the object class, property, and representation terms?  I've been
thinking about this, but still wrestling with the right ontology for their use
of "qualifiers" on these terms.

Thanks,
  Dave Carlson
  www.XMLmodeling.com


> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf
> Of Crowley, Rebecca
> Sent: Thursday, February 09, 2006 3:20 PM
> To: [hidden email]
> Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>
>
> Thomas:
>  
> I think I can interpret your question in two ways - and I am
> not sure which is the right way.
>  
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI
> has a ISO/IEC 111179 compliant registry that links Data
> Element Concepts, Valid Values, Representation Terms, and
> other administered components to concepts in a DL based
> ontology - NCI Thesaurus. NCI has made some extensions to the
> ISO/IEC standard - which I think they have proposed back to
> the standards organization for linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179
> compliant registries.
>  
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source
> except for the current vocabulary server.
>  
> Several of us (inside and outside NCICB) who worked on
> implementing a data standard (the CAP protocols)  using caDSR
> and NCI Thesaurus are just finishing a paper that describes
> the whole semantic linking thing. Once we are done with it -
> I'd be happy to forward it.
>  
> But... if what you are proposing is to actually implement the
> ISO standard as a frame system in protege - I think that
> would be very useful...I sort of half-heartedly attempted
> this myself last year but ran out of gas while reading
> through the standard :-) If that's what you mean -- Cecil
> Lynch might have done something like this...Cecil are you out there?
>  
> R Crowley
> University of Pittsburgh
>


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

dcarlson
In reply to this post by Crowley, Rebecca

Hi Rebecca,

I am also interested in taking a deeper look at your iso 11179 work.  My focus
is on the e-commerce domain, where several industry groups are using 11179
(parts 2 & 5, related to naming and classification) to standardize the naming of
reusable "Core Components" for orders, invoices, etc.

I took a quick look at your site, but have you created an ontology for
representing the object class, property, and representation terms?  I've been
thinking about this, but still wrestling with the right ontology for their use
of "qualifiers" on these terms.

Thanks,
  Dave Carlson
  www.XMLmodeling.com


> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf
> Of Crowley, Rebecca
> Sent: Thursday, February 09, 2006 3:20 PM
> To: [hidden email]
> Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>
>
> Thomas:
>  
> I think I can interpret your question in two ways - and I am
> not sure which is the right way.
>  
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI
> has a ISO/IEC 111179 compliant registry that links Data
> Element Concepts, Valid Values, Representation Terms, and
> other administered components to concepts in a DL based
> ontology - NCI Thesaurus. NCI has made some extensions to the
> ISO/IEC standard - which I think they have proposed back to
> the standards organization for linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179
> compliant registries.
>  
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source
> except for the current vocabulary server.
>  
> Several of us (inside and outside NCICB) who worked on
> implementing a data standard (the CAP protocols)  using caDSR
> and NCI Thesaurus are just finishing a paper that describes
> the whole semantic linking thing. Once we are done with it -
> I'd be happy to forward it.
>  
> But... if what you are proposing is to actually implement the
> ISO standard as a frame system in protege - I think that
> would be very useful...I sort of half-heartedly attempted
> this myself last year but ran out of gas while reading
> through the standard :-) If that's what you mean -- Cecil
> Lynch might have done something like this...Cecil are you out there?
>  
> R Crowley
> University of Pittsburgh
>
> ________________________________
>
> From: [hidden email] on behalf of Thomas
> Sent: Thu 2/9/2006 3:46 PM
> To: [hidden email]
> Subject: [protege-discussion] ISO 11179 and ontology modeling
>
>
>
>
> Dear Protege User(s),
>
> I'm currently working on a project which purpose is to
> implement an ISO 11179 compliant registry of standard
> data elements (in a pharmaceutical clinical
> environment).
>
> Based on what I know from the ISO 11179 standard,
> registered data elements could be classified in an
> ontological manner...
>
> Is anybody aware of a succesful implementation of such
> a standard, including ontological aspects ?
>
> Thanks in advance for your help.
>
> Kind Regards,
>
> Thomas
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>
>
>
> --
> Attached file removed by Ecartis and put at
> URL below --
> -- Type: application/ms-tnef
> -- Size: 5k (5702 bytes)
> -- URL :
> http://protege.stanford.edu/mail_archive/attac> hments/16-winmail.dat
>
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Elisa Kendall
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Hi Rebecca and Dave,

There is some work on the next version of 11179 ongoing, led by Lawrence
Berkeley Laboratories.  Kevin
Keck at LBL has created an ontology of 11179 Part 3, available from the
XMDR web site, at
http://hpcrd.lbl.gov/SDM/XMDR/overview.html.  I do not know how much
testing has been done using
this ontology, or what applications support it, but hope that it turns
out to be useful.  I'm sure that
Kevin would appreciate any feedback you have once you've reviewed this
as well.

Best regards,

Elisa


Dave Carlson wrote:

>Hi Rebecca,
>
>I am also interested in taking a deeper look at your iso 11179 work.  My focus
>is on the e-commerce domain, where several industry groups are using 11179
>(parts 2 & 5, related to naming and classification) to standardize the naming of
>reusable "Core Components" for orders, invoices, etc.
>
>I took a quick look at your site, but have you created an ontology for
>representing the object class, property, and representation terms?  I've been
>thinking about this, but still wrestling with the right ontology for their use
>of "qualifiers" on these terms.
>
>Thanks,
>  Dave Carlson
>  www.XMLmodeling.com
>
>
>  
>
>>-----Original Message-----
>>From: [hidden email]
>>[mailto:[hidden email]] On Behalf
>>Of Crowley, Rebecca
>>Sent: Thursday, February 09, 2006 3:20 PM
>>To: [hidden email]
>>Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>>
>>
>>Thomas:
>>
>>I think I can interpret your question in two ways - and I am
>>not sure which is the right way.
>>
>>The Cancer Bioinformatics Grid (caBIG) a project of the NCI
>>has a ISO/IEC 111179 compliant registry that links Data
>>Element Concepts, Valid Values, Representation Terms, and
>>other administered components to concepts in a DL based
>>ontology - NCI Thesaurus. NCI has made some extensions to the
>>ISO/IEC standard - which I think they have proposed back to
>>the standards organization for linking to an ontology in this
>>fashion -this is not currently required in ISO/IEC 11179
>>compliant registries.
>>
>>You can read more at:
>>http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
>>Most everything you would need that they use is open source
>>except for the current vocabulary server.
>>
>>Several of us (inside and outside NCICB) who worked on
>>implementing a data standard (the CAP protocols)  using caDSR
>>and NCI Thesaurus are just finishing a paper that describes
>>the whole semantic linking thing. Once we are done with it -
>>I'd be happy to forward it.
>>
>>But... if what you are proposing is to actually implement the
>>ISO standard as a frame system in protege - I think that
>>would be very useful...I sort of half-heartedly attempted
>>this myself last year but ran out of gas while reading
>>through the standard :-) If that's what you mean -- Cecil
>>Lynch might have done something like this...Cecil are you out there?
>>
>>R Crowley
>>University of Pittsburgh
>>
>>________________________________
>>
>>From: [hidden email] on behalf of Thomas
>>Sent: Thu 2/9/2006 3:46 PM
>>To: [hidden email]
>>Subject: [protege-discussion] ISO 11179 and ontology modeling
>>
>>
>>
>>
>>Dear Protege User(s),
>>
>>I'm currently working on a project which purpose is to
>>implement an ISO 11179 compliant registry of standard
>>data elements (in a pharmaceutical clinical
>>environment).
>>
>>Based on what I know from the ISO 11179 standard,
>>registered data elements could be classified in an
>>ontological manner...
>>
>>Is anybody aware of a succesful implementation of such
>>a standard, including ontological aspects ?
>>
>>Thanks in advance for your help.
>>
>>Kind Regards,
>>
>>Thomas
>>
>>--------------------------------------------------------------
>>-----------
>>To unsubscribe go to
>>http://protege.stanford.edu/community/subscrib> e.html
>>
>>
>>
>>
>>--
>>Attached file removed by Ecartis and put at
>>URL below --
>>-- Type: application/ms-tnef
>>-- Size: 5k (5702 bytes)
>>-- URL :
>>http://protege.stanford.edu/mail_archive/attac> hments/16-winmail.dat
>>
>>
>>--------------------------------------------------------------
>>-----------
>>To unsubscribe go to
>>http://protege.stanford.edu/community/subscrib> e.html
>>
>>
>>    
>>
>
>
>-------------------------------------------------------------------------
>To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
>
>  
>



-- Attached file removed by Ecartis and put at URL below --
-- Type: text/x-vcard
-- Size: 328 bytes
-- URL : http://protege.stanford.edu/mail_archive/attachments/02-ekendall.vcf


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Paul Prueitt
In reply to this post by dcarlson


Rebecca and Dave,

I have a background in modeling biological expression (initially
(dissertation 1989) with reaction networks and first order differential
equations (artificial neural networks - working with Grossberg), and in
theoretical immunology.

But over the past decade I have worked on ontological modeling using
semiotics and non standard logics.

(Sorry , I am trying to give context)


Recently I have been looking at a representation of the OASIS SOA-IM
(service oriented architecture information model) with its 38 "data objects"
as an RDF file (so that my programmers could use a standard visualization
interface using the RDF standard).  This is not easy at all.  I am stuck.

I have also been following the BioPAX work (as an outsider) since I am
interested in issues related to incomplete knowledge or knowledge that is
either not classified correctly or is understood incorrectly by the current
state of medical science.  Then there is the more difficult issue of
degeneracy in cellular or gene expression (as discussed by Edelman).

I am not aware of the ISO 11179  but am reading

http://metadata-standards.org/metadata-stds/Document-library/Documents-by-nu
mber/WG2-N0851-N0900/WG2-N0874-XMDR-Whitepaper-on-relationship-modeling-in-I
SOIEC-11179.htm

In the Protege OWL list group I am trying to see if a group might form (like
a working group) with the task of "translating" the SOA-IM (with is fairly
simple) to the simplest form of RDF possible; and then pushing that model in
the direction of OW*L ....  (the email is routed away from this discussion
if one mentions this word "OW*L... so sorry for this silliness)

or in the direction of Protege-Frames.  The other forum seems to not have
anyone who is interested in the translation working group concept.

The issues seem to be similar, whether one is talking about web services or
medical research, one needs to examine the assumptions made about what
"expressiveness" actually should mean.  What is a property, an attribute, a
class, etc...


The alternative to first order logics is not clear, this is true.

But at least there might be an agreement that there needs to be defined a
true and complete alternative?


I like some of what Alan Rector says about using OWL as an information
retrieval mechanism.  But the assertion of truth seems to be misplaced and
this can lead to side effects.


So I am open to suggestions.


Dr Paul S Prueitt
The Taos Research Institute
Taos New Mexico










-----Original Message-----
From: [hidden email]
[mailto:[hidden email]]On Behalf Of Dave
Carlson
Sent: Thursday, February 09, 2006 6:27 PM
To: [hidden email]
Subject: [protege-discussion] Re: ISO 11179 and ontology modeling



Hi Rebecca,

I am also interested in taking a deeper look at your iso 11179 work.  My
focus
is on the e-commerce domain, where several industry groups are using 11179
(parts 2 & 5, related to naming and classification) to standardize the
naming of
reusable "Core Components" for orders, invoices, etc.

I took a quick look at your site, but have you created an ontology for
representing the object class, property, and representation terms?  I've
been
thinking about this, but still wrestling with the right ontology for their
use
of "qualifiers" on these terms.

Thanks,
  Dave Carlson
  www.XMLmodeling.com


> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf
> Of Crowley, Rebecca
> Sent: Thursday, February 09, 2006 3:20 PM
> To: [hidden email]
> Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>
>
> Thomas:
>
> I think I can interpret your question in two ways - and I am
> not sure which is the right way.
>
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI
> has a ISO/IEC 111179 compliant registry that links Data
> Element Concepts, Valid Values, Representation Terms, and
> other administered components to concepts in a DL based
> ontology - NCI Thesaurus. NCI has made some extensions to the
> ISO/IEC standard - which I think they have proposed back to
> the standards organization for linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179
> compliant registries.
>
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source
> except for the current vocabulary server.
>
> Several of us (inside and outside NCICB) who worked on
> implementing a data standard (the CAP protocols)  using caDSR
> and NCI Thesaurus are just finishing a paper that describes
> the whole semantic linking thing. Once we are done with it -
> I'd be happy to forward it.
>
> But... if what you are proposing is to actually implement the
> ISO standard as a frame system in protege - I think that
> would be very useful...I sort of half-heartedly attempted
> this myself last year but ran out of gas while reading
> through the standard :-) If that's what you mean -- Cecil
> Lynch might have done something like this...Cecil are you out there?
>
> R Crowley
> University of Pittsburgh
>
> ________________________________
>
> From: [hidden email] on behalf of Thomas
> Sent: Thu 2/9/2006 3:46 PM
> To: [hidden email]
> Subject: [protege-discussion] ISO 11179 and ontology modeling
>
>
>
>
> Dear Protege User(s),
>
> I'm currently working on a project which purpose is to
> implement an ISO 11179 compliant registry of standard
> data elements (in a pharmaceutical clinical
> environment).
>
> Based on what I know from the ISO 11179 standard,
> registered data elements could be classified in an
> ontological manner...
>
> Is anybody aware of a succesful implementation of such
> a standard, including ontological aspects ?
>
> Thanks in advance for your help.
>
> Kind Regards,
>
> Thomas
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>
>
>
> --
> Attached file removed by Ecartis and put at
> URL below --
> -- Type: application/ms-tnef
> -- Size: 5k (5702 bytes)
> -- URL :
> http://protege.stanford.edu/mail_archive/attac> hments/16-winmail.dat
>
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html



-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Crowley, Rebecca
In reply to this post by Thomas-60
Dave:
 
No - not at all!!!
 
Actually the work that I was referring to in the first paragraph is work that NCICB did (a bunch of people there including Denise Warzel, Frank Hartel, George Komatsoulis and lots of others that I am undoubtedy leaving out.) What they have done is to implement ISO/IEC and to add to it a number of ways to link the administered components to an ontology. This is an extremely useful thing however, because it means that the data models themselves are entirely semantically annotated. And any terminology could potentially be used as the source of concepts. The more formal terminologies could support inference, aggregation, etc across the data model.
 
Several of us in the caBIG program used what those guys have developed and created a reasonably sized set of data elements that were linked to the ontology in this way. You might take a quick look at the CAP Cancer Checklist Models in the NCICB CDE Browser, and check out how they are linked to the NCI Thesaurus. This is the model that I worked on.
 
You are 100% correct that NCICB did not however create an ontology representing ISO/IEC 11179. As I said, I started and gave up on this while slogging through Parts 1-6 of that standard. At the time I did not fully understand ISO/IEC 11179 - so that might have been a big part of the problem. But I think it's a really interesting idea and I was sort of hoping that someone would try this and make it available to others
 
Hope that clarifies.
rebecca

________________________________

From: [hidden email] on behalf of Dave Carlson
Sent: Thu 2/9/2006 8:26 PM
To: [hidden email]
Subject: [protege-discussion] Re: ISO 11179 and ontology modeling




Hi Rebecca,

I am also interested in taking a deeper look at your iso 11179 work.  My focus
is on the e-commerce domain, where several industry groups are using 11179
(parts 2 & 5, related to naming and classification) to standardize the naming of
reusable "Core Components" for orders, invoices, etc.

I took a quick look at your site, but have you created an ontology for
representing the object class, property, and representation terms?  I've been
thinking about this, but still wrestling with the right ontology for their use
of "qualifiers" on these terms.

Thanks,
  Dave Carlson
  www.XMLmodeling.com


> -----Original Message-----
> From: [hidden email]
> [mailto:[hidden email]] On Behalf
> Of Crowley, Rebecca
> Sent: Thursday, February 09, 2006 3:20 PM
> To: [hidden email]
> Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>
>
> Thomas:
>
> I think I can interpret your question in two ways - and I am
> not sure which is the right way.
>
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI
> has a ISO/IEC 111179 compliant registry that links Data
> Element Concepts, Valid Values, Representation Terms, and
> other administered components to concepts in a DL based
> ontology - NCI Thesaurus. NCI has made some extensions to the
> ISO/IEC standard - which I think they have proposed back to
> the standards organization for linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179
> compliant registries.
>
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source
> except for the current vocabulary server.
>
> Several of us (inside and outside NCICB) who worked on
> implementing a data standard (the CAP protocols)  using caDSR
> and NCI Thesaurus are just finishing a paper that describes
> the whole semantic linking thing. Once we are done with it -
> I'd be happy to forward it.
>
> But... if what you are proposing is to actually implement the
> ISO standard as a frame system in protege - I think that
> would be very useful...I sort of half-heartedly attempted
> this myself last year but ran out of gas while reading
> through the standard :-) If that's what you mean -- Cecil
> Lynch might have done something like this...Cecil are you out there?
>
> R Crowley
> University of Pittsburgh
>
> ________________________________
>
> From: [hidden email] on behalf of Thomas
> Sent: Thu 2/9/2006 3:46 PM
> To: [hidden email]
> Subject: [protege-discussion] ISO 11179 and ontology modeling
>
>
>
>
> Dear Protege User(s),
>
> I'm currently working on a project which purpose is to
> implement an ISO 11179 compliant registry of standard
> data elements (in a pharmaceutical clinical
> environment).
>
> Based on what I know from the ISO 11179 standard,
> registered data elements could be classified in an
> ontological manner...
>
> Is anybody aware of a succesful implementation of such
> a standard, including ontological aspects ?
>
> Thanks in advance for your help.
>
> Kind Regards,
>
> Thomas
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>
>
>
> --
> Attached file removed by Ecartis and put at
> URL below --
> -- Type: application/ms-tnef
> -- Size: 5k (5702 bytes)
> -- URL :
> http://protege.stanford.edu/mail_archive/attac> hments/16-winmail.dat
>
>
> --------------------------------------------------------------
> -----------
> To unsubscribe go to
> http://protege.stanford.edu/community/subscrib> e.html
>
>


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html




-- Attached file removed by Ecartis and put at URL below --
-- Type: application/ms-tnef
-- Size: 9k (9309 bytes)
-- URL : http://protege.stanford.edu/mail_archive/attachments/17-winmail.dat


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

the frame paradigm

Paul Prueitt
In reply to this post by dcarlson



http://www.bcngroup.org/beadgames/generativeMethodology/162.htm


I post into this web log so that I can edit later and so that the formating
makes issues clearer.

I am interesting in taking this discussion off the Protege forum..

So perhaps those who might be interesting in collaboration could send an
email to me directly

[hidden email]



-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Thomas-60
In reply to this post by Crowley, Rebecca

Crowley, Rebecca <CrowleyRS <at> upmc.edu> writes:

>
> Thomas:
>
> I think I can interpret your question in two ways - and I am not sure which is
the right way.
>
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI has a ISO/IEC
111179 compliant registry that
> links Data Element Concepts, Valid Values, Representation Terms, and other
administered components to
> concepts in a DL based ontology - NCI Thesaurus. NCI has made some extensions
to the ISO/IEC standard -
> which I think they have proposed back to the standards organization for
linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179 compliant registries.
>
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source except for the
current vocabulary server.
>
> Several of us (inside and outside NCICB) who worked on implementing a data
standard (the CAP protocols)
> using caDSR and NCI Thesaurus are just finishing a paper that describes the
whole semantic linking thing.
> Once we are done with it - I'd be happy to forward it.
>
> But... if what you are proposing is to actually implement the ISO standard as
a frame system in protege - I
> think that would be very useful...I sort of half-heartedly attempted this
myself last year but ran out of
> gas while reading through the standard  If that's what you mean -- Cecil Lynch
might have done something

> like this...Cecil are you out there?
>
> R Crowley
> University of Pittsburgh
>
> ________________________________
>
> From: protege-discussion-bounce <at> SMI.Stanford.EDU on behalf of Thomas
> Sent: Thu 2/9/2006 3:46 PM
> To: protege-discussion <at> SMI.Stanford.EDU
> Subject: [protege-discussion] ISO 11179 and ontology modeling
>
> Dear Protege User(s),
>
> I'm currently working on a project which purpose is to
> implement an ISO 11179 compliant registry of standard
> data elements (in a pharmaceutical clinical
> environment).
>
> Based on what I know from the ISO 11179 standard,
> registered data elements could be classified in an
> ontological manner...
>
> Is anybody aware of a succesful implementation of such
> a standard, including ontological aspects ?
>
> Thanks in advance for your help.
>
> Kind Regards,
>
> Thomas
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>

Hi Rebecca,

Thanks for the feedback.

I had a (close) look at the caDSR and its linking with the EVS (NCI
Thesaurus/Metathesaurus)...Very impressive job...
I understand the semantic tagging capability as a very useful feature for
getting full control among administered items semantic.

The thing that remains unclear is the potential usage of ontology as formal
classification/conceptual schemes. If I understand the ISO 11179 well, this part
is meant to better "organize" administered items from a taxonomic to a full
ontological manner.

I'm little confused about how this ontological approach could be (has been?)
implemented...

Or did I completely miss the point ?

Sorry for those potential dummy questions, but fully understanding potential
ontological aspect of the ISO 11179 would definitely help me moving forward.

To answer your last question, I'm not proposing to represent the ISO as a frame
system in Protege. I'd rather think about building domain-specific (clinical
development) ontologies that I could attach to standardly defined data elements
(as common core vocabulary). Would this approach be realistic ?

Regards,

Thomas


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

Crowley, Rebecca
In reply to this post by Thomas-60
Thomas:
 
It's a big, big question!
I can tell you how it is being done in caBIG - and I think that they do it in a reasonable way, though I am certain that there are other reasonable ways too!
 
Clearly, the most obvious immediate advantage of the semantic tagging - is the potential for semantic interoperability. When caBIG models are loaded, Class_Attribute of each class are mapped to a CDE using their semantics. Regardless of what the class or attribute may be called - if the semantics align, with an existing CDE - then that CDE will be used by designation. Thus a model that has a class Agent with attribute NSC_number and another model that has a class Drug with attribute FDA number are going to be able to interoperate as long as they are using the same concepts within the DEC and they have the same representation. In this case, the model that registers second is going to end up using the same CDE as the one registered first - for which a new CDE was created.
 
Another fundamental thing to talk about  is just the idea of an intermediate representation between model and ontology. And why that is useful. This is the role that ISO/IEC 11179 really plays in caBIG.  Data models are annotated with ontologies via the intermediate ISO/IEC 11179 representation. So the ontology is truly attempting to be a canonical representation of knowledge across the entire domain, whilst we still have data models/metadata that are utterly bound by context. The intermediate layer gives you a place to make context specific assertions about entities, and to assert relationships which may not yet or ever exist in the ontology.
 
How does this play out in the models? Basically, the ontology stands alone in some sense. But the concepts in the ontology are used as tokens that annotate the models. Sometimes, the models are going to be in perfect synch with the ontology (check out the TNM Cancer Findings in my model) but sometimes they may be post-coordinated within the metadata (check out the SurgicalPathologySpecimen classes). So the idea is that you might have a class for homeworkMyDogAte which post-coordinates concepts within the ontology for 'homework '(Object Class) and 'dog' (OC Qualifier1) and 'ate' (OC qualifier2). Noone else but you cares about homework your dog ate. So homeworkMyDogAte is not going to be expressed as a concept within the vocabulary/ontology - it exists only within your metadata.
 
As you can immediately see, this leads to some complications with using the data models/ metadata for reasoning and classification. This might be the main price we pay for this intermediate layer. Because the semantics expressed  in the metadata may or may not be expressed within the DL formalism. In fact, to make it even worse, the metadata could well express relationships which are in direct contradiction to the ontology. I would argue that this is an essential requirement of our research oriented semantic modeling environment, and that it is a very useful feature of this middle layer. Researchers must be free to express in their data models hypotheses which contradict canonical knowledge.
 
(BTW, I say that the 'ontology stands alone in some sense' because the reality of the matter is that as we are developing these semantically annotated data models, we are working with the terminologists to figure out what has to be added or changed in the terminology)
 
You raised the interesting question of ISO classification schemes and what is there relationship to the ontology. In caBIG classification schemes are basically used as packages - we construct our UML models and the packages become classification schemes - used as data model specific methods for aggregating CDEs,  namespace etc. They are not used to create or reproduce ontologic taxonomies. What all of this means (I think) is that we are going to need to do all of our reasoning using the relationships expressed in the ontology. Therefore there will be parts of the metadata that we will not be able to easily reason about. I think that once you permit post-coordination in the metadata (with the qualifiers) as opposed to having every object class and property be a primitive - then you really make it very difficult to use the classification schemes in any other way. Though that might just be my biased opinion.
 
So far, few people in caBIG have attempted to use the ontology relationships to do anything within the data models/metadata. One potentially very useful thing to try would be to use the NCI taxonomies as a way to aggregate across models without any shared superclass, creating a kind of 'virtual superclass' based on semantics. However, to be honest, I think we will not really know much about how to make inferences using these semantic models until a few of us have tried to do something other than interoperate.
 
Hope that helps to at least clarify how we are doing it. It's been an awfully difficult thing for me to wrap my head around over the last year or two, so I know how you feel.
 
Best regards,
rebecca
 
-----------------------------
 
Hi Rebecca,

Thanks for the feedback.

I had a (close) look at the caDSR and its linking with the EVS (NCI
Thesaurus/Metathesaurus)...Very impressive job...
I understand the semantic tagging capability as a very useful feature for
getting full control among administered items semantic.

The thing that remains unclear is the potential usage of ontology as formal
classification/conceptual schemes. If I understand the ISO 11179 well, this part
is meant to better "organize" administered items from a taxonomic to a full
ontological manner.

I'm little confused about how this ontological approach could be (has been?)
implemented...

Or did I completely miss the point ?

Sorry for those potential dummy questions, but fully understanding potential
ontological aspect of the ISO 11179 would definitely help me moving forward.

To answer your last question, I'm not proposing to represent the ISO as a frame
system in Protege. I'd rather think about building domain-specific (clinical
development) ontologies that I could attach to standardly defined data elements
(as common core vocabulary). Would this approach be realistic ?

Regards,

Thomas


________________________________

From: [hidden email] on behalf of Thomas
Sent: Fri 2/10/2006 3:58 PM
To: [hidden email]
Subject: [protege-discussion] Re: ISO 11179 and ontology modeling




Crowley, Rebecca <CrowleyRS <at> upmc.edu> writes:

>
> Thomas:
>
> I think I can interpret your question in two ways - and I am not sure which is
the right way.
>
> The Cancer Bioinformatics Grid (caBIG) a project of the NCI has a ISO/IEC
111179 compliant registry that
> links Data Element Concepts, Valid Values, Representation Terms, and other
administered components to
> concepts in a DL based ontology - NCI Thesaurus. NCI has made some extensions
to the ISO/IEC standard -
> which I think they have proposed back to the standards organization for
linking to an ontology in this
> fashion -this is not currently required in ISO/IEC 11179 compliant registries.
>
> You can read more at:
> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
> Most everything you would need that they use is open source except for the
current vocabulary server.
>
> Several of us (inside and outside NCICB) who worked on implementing a data
standard (the CAP protocols)
> using caDSR and NCI Thesaurus are just finishing a paper that describes the
whole semantic linking thing.
> Once we are done with it - I'd be happy to forward it.
>
> But... if what you are proposing is to actually implement the ISO standard as
a frame system in protege - I
> think that would be very useful...I sort of half-heartedly attempted this
myself last year but ran out of
> gas while reading through the standard  If that's what you mean -- Cecil Lynch
might have done something

> like this...Cecil are you out there?
>
> R Crowley
> University of Pittsburgh
>
> ________________________________
>
> From: protege-discussion-bounce <at> SMI.Stanford.EDU on behalf of Thomas
> Sent: Thu 2/9/2006 3:46 PM
> To: protege-discussion <at> SMI.Stanford.EDU
> Subject: [protege-discussion] ISO 11179 and ontology modeling
>
> Dear Protege User(s),
>
> I'm currently working on a project which purpose is to
> implement an ISO 11179 compliant registry of standard
> data elements (in a pharmaceutical clinical
> environment).
>
> Based on what I know from the ISO 11179 standard,
> registered data elements could be classified in an
> ontological manner...
>
> Is anybody aware of a succesful implementation of such
> a standard, including ontological aspects ?
>
> Thanks in advance for your help.
>
> Kind Regards,
>
> Thomas
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html




-- Attached file removed by Ecartis and put at URL below --
-- Type: application/ms-tnef
-- Size: 11k (11693 bytes)
-- URL : http://protege.stanford.edu/mail_archive/attachments/18-winmail.dat


-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html

Reply | Threaded
Open this post in threaded view
|

Re: ISO 11179 and ontology modeling

libren
In reply to this post by Thomas-60

> Clearly, the most obvious immediate advantage of the semantic tagging - is

> the potential for semantic interoperability.
Are there well-known researches on the subject some of you could point to ?
I am working on information retrieval based on activities context; contexts
are mapped into ontologies and have to interoperate with indexing
ontologies.

----- Original Message -----
From: "Crowley, Rebecca" <[hidden email]>
To: <[hidden email]>
Sent: Monday, February 13, 2006 3:36 AM
Subject: [protege-discussion] Re: ISO 11179 and ontology modeling


> Thomas:
>
> It's a big, big question!
> I can tell you how it is being done in caBIG - and I think that they do it

> in a reasonable way, though I am certain that there are other reasonable
> ways too!
>
> Clearly, the most obvious immediate advantage of the semantic tagging - is

> the potential for semantic interoperability. When caBIG models are loaded,

> Class_Attribute of each class are mapped to a CDE using their semantics.
> Regardless of what the class or attribute may be called - if the semantics

> align, with an existing CDE - then that CDE will be used by designation.
> Thus a model that has a class Agent with attribute NSC_number and another
> model that has a class D r u g with attribute FDA number are going to be
> able to interoperate as long as they are using the same concepts within
> the DEC and they have the same representation. In this case, the model
> that registers second is going to end up using the same CDE as the one
> registered first - for which a new CDE was created.
>
> Another fundamental thing to talk about  is just the idea of an
> intermediate representation between model and ontology. And why that is
> useful. This is the role that ISO/IEC 11179 really plays in caBIG.  Data
> models are annotated with ontologies via the intermediate ISO/IEC 11179
> representation. So the ontology is truly attempting to be a canonical
> representation of knowledge across the entire domain, whilst we still have

> data models/metadata that are utterly bound by context. The intermediate
> layer gives you a place to make context specific assertions about
> entities, and to assert relationships which may not yet or ever exist in
> the ontology.
>
> How does this play out in the models? Basically, the ontology stands alone

> in some sense. But the concepts in the ontology are used as tokens that
> annotate the models. Sometimes, the models are going to be in perfect
> synch with the ontology (check out the TNM Cancer Findings in my model)
> but sometimes they may be post-coordinated within the metadata (check out
> the SurgicalPathologySpecimen classes). So the idea is that you might have

> a class for homeworkMyDogAte which post-coordinates concepts within the
> ontology for 'homework '(Object Class) and 'dog' (OC Qualifier1) and 'ate'

> (OC qualifier2). Noone else but you cares about homework your dog ate. So
> homeworkMyDogAte is not going to be expressed as a concept within the
> vocabulary/ontology - it exists only within your metadata.
>
> As you can immediately see, this leads to some complications with using
> the data models/ metadata for reasoning and classification. This might be
> the main price we pay for this intermediate layer. Because the semantics
> expressed  in the metadata may or may not be expressed within the DL
> formalism. In fact, to make it even worse, the metadata could well express

> relationships which are in direct contradiction to the ontology. I would
> argue that this is an essential requirement of our research oriented
> semantic modeling environment, and that it is a very useful feature of
> this middle layer. Researchers must be free to express in their data
> models hypotheses which contradict canonical knowledge.
>
> (BTW, I say that the 'ontology stands alone in some sense' because the
> reality of the matter is that as we are developing these semantically
> annotated data models, we are working with the terminologists to figure
> out what has to be added or changed in the terminology)
>
> You raised the interesting question of ISO classification schemes and what

> is there relationship to the ontology. In caBIG classification schemes are

> basically used as packages - we construct our UML models and the packages
> become classification schemes - used as data model specific methods for
> aggregating CDEs,  namespace etc. They are not used to create or reproduce

> ontologic taxonomies. What all of this means (I think) is that we are
> going to need to do all of our reasoning using the relationships expressed

> in the ontology. Therefore there will be parts of the metadata that we
> will not be able to easily reason about. I think that once you permit
> post-coordination in the metadata (with the qualifiers) as opposed to
> having every object class and property be a primitive - then you really
> make it very difficult to use the classification schemes in any other way.

> Though that might just be my biased opinion.
>
> So far, few people in caBIG have attempted to use the ontology
> relationships to do anything within the data models/metadata. One
> potentially very useful thing to try would be to use the NCI taxonomies as

> a way to aggregate across models without any shared superclass, creating a

> kind of 'virtual superclass' based on semantics. However, to be honest, I
> think we will not really know much about how to make inferences using
> these semantic models until a few of us have tried to do something other
> than interoperate.
>
> Hope that helps to at least clarify how we are doing it. It's been an
> awfully difficult thing for me to wrap my head around over the last year
> or two, so I know how you feel.
>
> Best regards,
> rebecca
>
> -----------------------------
>
> Hi Rebecca,
>
> Thanks for the feedback.
>
> I had a (close) look at the caDSR and its linking with the EVS (NCI
> Thesaurus/Metathesaurus)...Very impressive job...
> I understand the semantic tagging capability as a very useful feature for
> getting full control among administered items semantic.
>
> The thing that remains unclear is the potential usage of ontology as
> formal
> classification/conceptual schemes. If I understand the ISO 11179 well,
> this part
> is meant to better "organize" administered items from a taxonomic to a
> full
> ontological manner.
>
> I'm little confused about how this ontological approach could be (has
> been?)
> implemented...
>
> Or did I completely miss the point ?
>
> Sorry for those potential dummy questions, but fully understanding
> potential
> ontological aspect of the ISO 11179 would definitely help me moving
> forward.
>
> To answer your last question, I'm not proposing to represent the ISO as a
> frame
> system in Protege. I'd rather think about building domain-specific
> (clinical
> development) ontologies that I could attach to standardly defined data
> elements
> (as common core vocabulary). Would this approach be realistic ?
>
> Regards,
>
> Thomas
>
>
> ________________________________
>
> From: [hidden email] on behalf of Thomas
> Sent: Fri 2/10/2006 3:58 PM
> To: [hidden email]
> Subject: [protege-discussion] Re: ISO 11179 and ontology modeling
>
>
>
>
> Crowley, Rebecca <CrowleyRS <at> upmc.edu> writes:
>
>>
>> Thomas:
>>
>> I think I can interpret your question in two ways - and I am not sure
>> which is
> the right way.
>>
>> The Cancer Bioinformatics Grid (caBIG) a project of the NCI has a ISO/IEC
> 111179 compliant registry that
>> links Data Element Concepts, Valid Values, Representation Terms, and
>> other
> administered components to
>> concepts in a DL based ontology - NCI Thesaurus. NCI has made some
>> extensions
> to the ISO/IEC standard -
>> which I think they have proposed back to the standards organization for
> linking to an ontology in this
>> fashion -this is not currently required in ISO/IEC 11179 compliant
>> registries.
>>
>> You can read more at:
>> http://ncicb.nci.nih.gov/NCICB/infrastructure/cacore_overview/cadsr
>> Most everything you would need that they use is open source except for
>> the
> current vocabulary server.
>>
>> Several of us (inside and outside NCICB) who worked on implementing a
>> data
> standard (the CAP protocols)
>> using caDSR and NCI Thesaurus are just finishing a paper that describes
>> the
> whole semantic linking thing.
>> Once we are done with it - I'd be happy to forward it.
>>
>> But... if what you are proposing is to actually implement the ISO
>> standard as
> a frame system in protege - I
>> think that would be very useful...I sort of half-heartedly attempted this
> myself last year but ran out of
>> gas while reading through the standard  If that's what you mean -- Cecil
>> Lynch
> might have done something
>> like this...Cecil are you out there?
>>
>> R Crowley
>> University of Pittsburgh
>>
>> ________________________________
>>
>> From: protege-discussion-bounce <at> SMI.Stanford.EDU on behalf of Thomas
>> Sent: Thu 2/9/2006 3:46 PM
>> To: protege-discussion <at> SMI.Stanford.EDU
>> Subject: [protege-discussion] ISO 11179 and ontology modeling
>>
>> Dear Protege User(s),
>>
>> I'm currently working on a project which purpose is to
>> implement an ISO 11179 compliant registry of standard
>> data elements (in a pharmaceutical clinical
>> environment).
>>
>> Based on what I know from the ISO 11179 standard,
>> registered data elements could be classified in an
>> ontological manner...
>>
>> Is anybody aware of a succesful implementation of such
>> a standard, including ontological aspects ?
>>
>> Thanks in advance for your help.
>>
>> Kind Regards,
>>
>> Thomas
>>
>> -------------------------------------------------------------------------
>> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>
>
>
>
> -- Attached file removed by Ecartis and put at URL below --
> -- Type: application/ms-tnef
> -- Size: 11k (11693 bytes)
> -- URL :
> http://protege.stanford.edu/mail_archive/attachments/18-winmail.dat
>
>
> -------------------------------------------------------------------------
> To unsubscribe go to http://protege.stanford.edu/community/subscribe.html
>

-------------------------------------------------------------------------
To unsubscribe go to http://protege.stanford.edu/community/subscribe.html