ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] entropy

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "Ravinder Thakur" <ravinderthakur@xxxxxxxxx>
Date: Thu, 30 Aug 2007 22:33:02 +0530
Message-id: <617073f10708301003j641fce4p479fa3fdb0eb0dea@xxxxxxxxxxxxxx>
thanks for your response. i also felt like there is no way to move forward with it. but another important point is wether this is a significant issue for ontologies in the long run. is this a potential problem in realizing the full potential of ontologies in perticular and semantic web in general ?
 
 
Your thought on this will be appriciated.


 
On 8/30/07, Pat Hayes <phayes@xxxxxxx> wrote:
>Hi Deb
>
>Thanks for forwarding Ravinders question.
>You mean its similar to my question about how to go about adding new
>dimensions to the framework?
>
>The suggestion of adding 'entropy' to the framework comes from the
>discussions around which there seems to be some agreement, that
>ontologies should  model 'change' - from simple change of usage of
>words and terminologies during time, to the more complex changes of
>states of entities, to the more ambitious modelling of 'unknown
>variables'
>briefly touched upon in a discussion with Ed (have a generic class of
>'unpredictable' random variables in any ontological model to represent
>at least symbolically 'uncertainty')
>
>Entropy seems to abstract quite well the principle of change,

Sorry, I strongly disagree. Change and entropy
are completely different ideas. They are related
by the thermodynamic theorems, but those apply
only in carefully controlled conditions (closed
systems, etc.) and are not a good basis for
general ontology work (for example, they do not
apply to the Internet as a whole, which consumes
vast amounts of energy). In any case there are
many kinds of change which are not related to
thermodynamics. Change is related to time, not
entropy. For a (now quite old) survey, see
http://www.ihmc.us/users/phayes/TimeCatalog.pdf

>without
>making any prescription at this stage on how change should be
>modelled

Good idea, although one should be aware that this
is one topic on which ontologists have probably
spent more effort and thought than any other, so
it is hardly a blank slate.

>, which probably depends on the domain being addressed.
>The current (hm, yesterday morning asia pacific time) entry on
>wikipedia gives a good enough introduction, with a clear reference to
>E in information theory
>
> http://en.wikipedia.org/wiki/Entropy
>
>The less organzeds is a system the higher its entropy - lots of
>changes going on in the web
>
>
>So maybe I was thinking of 2 questions
>
>1.is entropy a suitable dimension to be added to the framework to
>summarize the discourse on change

No.

>2. how do we 'change' the framework over time? (that matches both 1,
>as well as Ravindra's problem - which maybe others can address
>separately)

I don't think Ravindra's problem is directly
related to change at all: he is more concerned
with the multitude of ontologies all of which
conceptualise things differently and to which he
wants to apply global queries. He is simply out
of luck: what he wants to do (using SPARQL) is
not feasible in the current state of the art.

Pat

>Cheers
>
>PDM
>
>
>
>On 8/29/07, Deborah MacPherson <debmacp@xxxxxxxxx> wrote:
>>  Hi Paola,
>>
>>  Please see the message below forwarded from Ravinder Thakur on
>>  Linking-open-data at MIT. Will his problem be addressed in the framework?
>>  Does anyone have suggestions for Mr. Thakur and his collaborators at linking
>>  open data?
>>
>>   Thank you,
>>
>>   Deborah
>>
>>  Deborah L. MacPherson
>>  Projects Director, Accuracy&Aesthetics
>>  Specifier, WDG Architecture PLLC
>>
>>  ---------- Forwarded message ----------
>>  From: "Ravinder Thakur" < ravinderthakur@xxxxxxxxx>
>>  To: "Linking Open Data" < linking-open-data@xxxxxxxxxxxxxx >
>>  Date: Sun, 26 Aug 2007 22:18:09 -0700
>>  Subject: [Linking-open-data] RDF/SPARQL and consistent metadata about
>>  content/structure of ontologies
>>
>>   Hi,
>>
>>  For the past few days i have been trying to look at RDF/SPARQL from query
>>  designing point of view. What basically i have been trying to do is to take
>>  very simple /user/ queries (" e.g. Bill Gates net worth") and try to see if i
>>  can automatically create the corresponding queriers in SPARQL. The versatile
>>  nature of domain of data stored in RDF means that data from every domain
>>  would have its own ontology. And the knowledge/understanding of ontology is
>>  very much prerequisite to understanding the data based on that particular
>>  ontology.
>>
>>  This seems a little bit of problem to me. This means that we cannot build
>  > queries until we know that structure of ontologies and since there might be
>>  thousands of ontologies how are we supposed to know about all the
>>  ontologies. Moreover everyday new ontologies will be added to the web.? IS
>>  the assumption that the dust will settle after some time and we will have
>>  only very few good ontologies about the world ? This seems highly
>>  restrictive to me.
>>
>>  Obviously I want to write a query processor that is independent on the
>>  ontology that the data i am looking on is based upon. I don't want to know
>>  pre hand what ontology might correspond to the data i am looking . What i am
>>  looking for some consistent way to represent the /METADATA/ about the
>>  structure and content about ontologies so that from the metadata about the
>>  ontology, i can automatically know about the structure and relevance of the
>>  ontology for the query i have in hand.
>>
>>  Has anybody faced the problem i have highlighted ? or what i am saying is
>>  stupid :) and there is some simple way out there ?
>>
>>   Thanks
>>  Ravinder Thakur
>>
>>
>>  On 8/27/07, paola.dimaio@xxxxxxxxx < paola.dimaio@xxxxxxxxx> wrote:
>>  >
>>  > When is the next update of the framework due?
>>  > I propose we add 'entropy' as a new dimension
>>  >
>>  > Do I have to write something up? Or are the discussions so far on the
>>  > need to model
>>  > 'change' sufficient?
>>  >
>>  >
>>  > --
>>  > 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
>>  >
>>  >
>>
>>   On 8/27/07, paola.dimaio@xxxxxxxxx <paola.dimaio@xxxxxxxxx > wrote:
>>  >
>>  > When is the next update of the framework due?
>>  > I propose we add 'entropy' as a new dimension
>>  >
>>  > Do I have to write something up? Or are the discussions so far on the
>>  > need to model
>>  > 'change' sufficient?
>>  >
>>  >
>>  > --
>>  > 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
>>  >
>>  >
>>
>>
>>
>>  --
>>  *************************************************
>>  Deborah L. MacPherson
>>  Projects Director, Accuracy&Aesthetics
>>  Specifier, WDG Architecture PLLC
>>
>>  The content of this email may contain private
>>   and confidential information. Do not forward,
>>  copy, share, or otherwise distribute without
>>  explicit written permission from all
>>  correspondents.
>>
>>  **************************************************
>>
>>  _________________________________________________________________
>>  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
>


--
---------------------------------------------------------------------
IHMC            (850)434 8903 or (650)494 3973   home
40 South Alcaniz St.    (850)202 4416   office
Pensacola                       (850)202 4440   fax
FL 32502                        (850)291 0667    cell
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes


_________________________________________________________________
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



_________________________________________________________________
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)

<Prev in Thread] Current Thread [Next in Thread>