Re: [ontolog-forum] what is open ontology?

["Alexander Garcia Castro" <alexgarciac@xxxxxxxxx>]

I can understand the importance of knowing the sources of information that were used to develop the ontology. Shouldn't these be part of the annotation of the ontology? For instance, when developing ontologies one source of information are human experts, they give evidence and supporting material –books, papers, white papers, use cases, personal representations of the "this is what I mean".  But classes in an ontology may have, and this is usually the case, more than one source of information, and the source, as I have already mentioned, may come in different formats. Shouldn't these sources of information as well as their corresponding representations be part of the annotation for classes and properties?  By the same token one would have a summary for the whole ontology, have human experts been involved? Is this an ontology derived from texts? Is there any software using the ontology? If so is there a URL for this software?

 

This may facilitate the actual evaluation process, because will make it easier for the conceptual verification of the ontology, either by one external expert or by a group of experts. This may also facilitate the evolution of the ontology, as well as its reusability, because in this way others, who were not involved in the development of the ontology, may find it easier to understand the rational behind the actual structure of the is_a hierarchy, and the whys for the object properties.

On Sat, Apr 26, 2008 at 12:39 AM,  <paola.dimaio@xxxxxxxxx> wrote:

Wacek,

I am interested in the definitions and various interpretations of
'source' vs 'source code' or other that people come up with

However, in some of my earlier write ups on the notion of OO (open
ontology), I say that 'source' should be declared, I mean 'source of
knowledge', which is an additional  type of 'source' that I think
should be addressed for the purpose of this discussion

The source of knowledge is very important for ontology. Say for
example  in a medical ontology, ontological facts are derived from a
given 'belief systems' and according to the accepted BOK (body of
knowledge) and or text books that are considered 'the bible' in that
domain, by a given school of thought. If we compile a medical ontology
taking into account chinese medicine textbooks, we end up with an
entirely different ontology, and so forth.

So when presented with a 'medical ontology', I would like to know from
which source of knowledge the ontological facts are derived from, so
that I can put them into the appropriate 'perspective'

Same applies to any ontology. This would enable the user to validate,
verify, compare, cross reference etc, ontological assertions,  instead
of expecting  'usesr' to take the ontology at face value and 'as
given'.

Paola DM

On Sat, Apr 26, 2008 at 3:00 AM, Wacek Kusnierczyk
<Waclaw.Marcin.Kusnierczyk@xxxxxxxxxxx> wrote:
> John F. Sowa wrote:
>  > Pat,
>  >
>  > I would claim that an ontology could be open in the
>  > same sense that programs can be open.
>  >
>  > PH> I don't think it makes sense to refer to an ontology
>  >  > as being open (or not).
>  >
>  > Although an ontology is not likely to be compiled to
>  > executable machine code, it can be converted to internal
>  > representations that are humanly unreadable.  For example,
>  > some would regard OWL as having that property.  But OWL can
>  > be translated back to more readable notations, if desired.
>  >
>  > However, anyone who wanted to keep the knowledge representation
>  > proprietary -- including the ontology, axioms, etc. -- could
>  > certainly do so by compiling them to some proprietary form.
>  >
>  > There could be many reasons for doing such a compilation.
>  > RDF and OWL, for example, are extremely verbose, and many
>  > systems compile them to more efficient forms for processing.
>  > If those internal forms are kept proprietary, the effect
>  > would be to keep the ontology as "closed" as any program.
>  >
>  >
>  There is also the issue of what is to be understood as 'source'.  An
>  ontology in, e.g., OWL can be considered as the result of a compilation
>  from sources such as experimental evidence or prior expert knowledge.
>  Now, if you can get (and read) only the OWL encoding of the ontology,
>  but not the sources used during the compilation, it is obviously not
>  open source even though the final code is accessible.
>
>  Contrarily, if 'source' is to mean the final encoding (e.g., an OWL
>  file), then all software you can run is open source, since its source (=
>  the binaries according to this view) is readable, otherwise you would
>  not be able to run it.  So it would not make sense to call any software
>  'open source' at all.
>
>  vQ
>
>
>
>  _________________________________________________________________
>  Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
>  Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
>  Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
>  Shared Files: http://ontolog.cim3.net/file/
>  Community Wiki: http://ontolog.cim3.net/wiki/
>  To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
>
>

--

Paola Di Maio
School of IT
www.mfu.ac.th
*********************************************

_________________________________________________________________

Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx

-- 
Alexander Garcia
http://www.usefilm.com/photographer/75943.html
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (01)