W3C home > Mailing lists > Public > semantic-web@w3.org > January 2009

Re: Federated Ontology System: was semantic technologies training/request

From: Hugh Glaser <hg@ecs.soton.ac.uk>
Date: Wed, 7 Jan 2009 11:59:40 +0000
To: Azamat <abdoul@cytanet.com.cy>
CC: SW-forum <semantic-web@w3.org>
Message-ID: <C58A49AC.289CD%hg@ecs.soton.ac.uk>
Thanks Azamat.
I am trying to stay quite specific here.
You are making some statements about how you think the Semantic Web
should/will work.
You are supporting these statements by making assertions about other
systems, that may have relevance, but it is not always clear that you
establish what the relevance is.
And you seem to be taking specific views of those other systems that would
not always win 100% agreement with experts in those fields.
So I fail to see the value of your assertion that ”Like it or not, to order
and unify things, you need a central authority”, as you seem to make no
arguments to support it that actually refer to the domain we are discussing.

On 07/01/2009 11:43, "Azamat" <abdoul@cytanet.com.cy> wrote:

> On Wednesday, January 07, 2009 4:00 AM, Hugh wrote:
> "We often reflect our political beliefs in our technical work; it doesn't
>  validate either. And if you were a Syndicalist Anarchist, you might see the
> Semantic Web
> differently - but would that make your view any more correct?"
>
> Dear Hugh,
> It is funny how intelligent people may misread each other.
> This is not the matter of your politics. It might be surprising, but the
> words as notions tend to change their meanings, extending or narrowing their
> senses, either by the processes of generalization or specialization, or by
> spreading, or transference, or hyperbolizing, or by figurative extension.
> Now, dealing with the terms 'to unite' or 'to unify' [web resources], one
> have many connotations, as 'amalgamate, blend, cement, cohere, consolidate,
> combine, compound, connect, federate, integrate, link, merge, fuse, join',
> etc, but all with specific differences in meanings. What is good with
> 'federate', it suggests 'the act of unifying of heterogenous and autonomous
> entities under a central authority', which can be naturally extended as a
> 'central ontology".
> Keeping in mind its political and data base meanings, we can talk about a
> top-bottom globally federated ontology, having a flexible hierarchical
> network topology; the star, the mesh, the loop, the bar are all their.
> I have an impression that Enrico's NeOn scheme is mainly about the mesh
> ontology, a network their the component ontologies are directly to be
> connected to every other element ontologies.That's just impossible in all
> senses.
> All the world works like this, any real entities are interconnected via the
> fundamental elements, principles and laws; in the physical world, it is
> matter and energy, fundamental forces and interactions, gravitation,
> electromagnetism, weak and strong forces.
And to address the wider issue, which I know I should resist, but can't:
I really don't see that referring to these principles and laws helps to
support the statement you make in your next paragraph summing up, saying we
need a "central authority". I think electromagnetism seems to work pretty
well without an obvious central authority.
>
> Summing:
> Like it or not, to order and unify things, you need a central authority, in
> the Globally Federated Ontology (GFO), it is a global scheme of all
> resources, capable to integrate all sorts of particular schemes, data
> models, and views. This most challenging problem is be the first focus of
> any large-scale SW research projects, imho.
> Regards,
I'm not saying I think you are wrong, but I would be pleased to see
arguments directly addressing the technological issues.
Regards
Hugh
> Azamat Abdoullaev
> http://www.eis.com.cy

> [see how the idea of knowledge federation may work for sciences and
> technology as well].
>
> ----- Original Message -----
> From: "Hugh Glaser" <hg@ecs.soton.ac.uk>
> To: "Azamat" <abdoul@cytanet.com.cy>; "Enrico Motta" <e.motta@open.ac.uk>
> Cc: "SW-forum" <semantic-web@w3.org>
> Sent: Wednesday, January 07, 2009 4:00 AM
> Subject: Re: Federated Ontology System: was semantic technologies
> training/request
>
>
>> One thing which seemed a poor argument here:
>>
>>
>> On 06/01/2009 22:28, "Azamat" <abdoul@cytanet.com.cy> wrote:
>>
>>>
>>>
>>> On Tuesday, January 06, 2009 7:06 PM, Enrico Motta wrote:
>>> "...if you look at the SW as it is today, you can already see thousands
>>> and
>>> thousands of alternative
>>> conceptualizations. This is why the NeOn project is developing a variety
>>> of
>>> practical solutions
>>> that focus on enabling the development, maintenance and use of networked
>>> ontologies, rather than assuming that some individual or
>>> organization will give us the 'global ontology'."
>>>
>>> Dear Enrico,
>>>
>>> There is the whole point which somebody of us missing. The issue is, how
>>> you
>>> are doing the networking of domain ontologies.  What one read as a
>>> definition: "A Network of Ontologies is a collection of ontologies
>>> related
>>> together via a variety of different relationships such as mapping,
>>> modularization, version and  dependency relationships".  Softly speking,
>>> it
>>> is not very productive to think that way; for nobody in the world, in no
>>> time and money, is capable to interrelate in such ways an innumerable
>>> number
>>> of particular ontologies, distributed, autonomous and heterogeneous, with
>>> their specific local schemas, semantics, languages, formats, data models,
>>> and structures.
>>> Seemingly, You had an impression that i suggested a centralized, unitary
>>> ontology system governed by a single global schema. This is not the case.
>>> This matter is now also discussed on the Ontolog Forum. Some main points.
>>> I
>>> am talking about a realistic, flexible and scalable model of a federated
>>> (web) ontology [used in Ontopaedia, check the Index page,
>>> http://www.eis.com.cy]. The model implies such effective notions as
>>> "ontology federation", "federated ontology system", "federated global
>>> schema", 'federated ontology architecture", and "federated local
>>> schemas".
>>>  The notion of a federal union proved its viability in politics as a
>>> federal
>>> form of government, where power is divided between a central authority
>>> and
>>> regional authorities. Also, it was successfully applied in the database
>>> theory and practice, as "a federated architecture for database systems"
>>> or
>>> "a federated architecture for information management".
>> Notwithstanding discussions about which political systems have proved
>> themselves effective...
>> I really don't think anyone can claim that database federation has been
>> successful in the modern, open, web-enabled world.
>> The difficulty that I have in publishing my database and have it easily
>> federate with other databases is a (the?) major motivator for SW work.
>>> Now, alike with the power, knowledge is divided between a central
>>> ontology
>>> and regional ontologies. Then a federation ontology will consist of a
>>> single
>>> central ontology (maintaining the global schema, the semantics, the
>>> topology, the entry of new ontologies) and a multitude of component
>>> ontologies with own
>>> local schemas,  but members of the federation. There are technical
>>> issues,
>>> such as federated mechanism, semantic management, schemas integration and
>>> coordination, search, information retrieval and query processing, etc.
>>> But
>>> what is essential: the reality of the concept of Federated Ontology
>>> System,
>>> which, to my experience, looks more promising than any (botom-top)
>>> nonfederated ontology systems, either unitary or centralized or loose and
>>> unconnected, currently prevailing and propagating as pandemic on the WWW.
>> We often reflect our political beliefs in our technical work; it doesn't
>> validate either.
>> And if you were a Syndicalist Anarchist, you might see the Semantic Web
>> differently - but would that make your view any more correct?
>> Best
>> Hugh
>>>
>>> kind regards,
>>> Azamat Abdoullaev
>>> http://www.eis.com.cy

>>>
>>>
>>> ----- Original Message -----
>>> From: "Enrico Motta" <e.motta@open.ac.uk>
>>> To: "Azamat" <abdoul@cytanet.com.cy>
>>> Cc: "'SW-forum'" <semantic-web@w3.org>; <paola.dimaio@gmail.com>
>>> Sent: Tuesday, January 06, 2009 7:06 PM
>>> Subject: Re: semantic technologies training/request
>>>
>>>
>>>
>>> At 20:18 +0200 30/12/08, Azamat wrote:
>>>> €
>>>> On Sunday, December 28, 2008 12:55, Paola wrote:
>>>> "PMI am starting to be introduced to great sw tools being released by
>>>> the
>>>> various EU funded projects, for which lots and lots
>>>> of public money is been used such as
>>>> <http://ontoware.org/>http://ontoware.org/ as well as lots of others"
>>>>
>>>> Paola,
>>>> Thanks for an intersting link.
>>>>
>>>> I was intrigued to see what is presented as "ontoware", finding the
>>>> following  project as most engaging,
>>>> <http://www.neon-project.org/web-content/>http://www.neon-project.org/web-c

>>>> on
>>>> tent/.
>>>> Being surprised with overwhelming ontological activities, one is
>>>> attracted
>>>> to browse the project boasting that:
>>>> "NeOn is a 14.7 million Euros project involving 14 European partners and
>>>> co-funded by the European Commission’s Sixth Framework Programme under
>>>> grant number IST-2005-027595. NeOn started in March 2006 and has a
>>>> duration
>>>> of 4 years. Our aim is to advance the state of the art in using
>>>> ontologies
>>>> for large-scale semantic applications in the distributed organizations.
>>>> Particularly, we aim at improving the capability to handle multiple
>>>> networked ontologies that exist in a particular context, are created
>>>> collaboratively, and might be highly dynamic and constantly evolving."
>>>>
>>>> Here is the NeOn basic defintion: "A Network of Ontologies is a
>>>> collection
>>>> of ontologies related together via a variety of different relationships
>>>> such as mapping, modularization, version and  dependency relationships".
>>>> Indeed, all fundamental troubles are in assumptions and presumptions.
>>>>
>>> Glancing at the content, one might start
>>> questioning the promised tools and applications
>>> for justified reasons. First, instead of a
>>> variety of diverse, modular, individual
>>> ontologies, the Semantic Web implies an
>>> integrated collection of domain ontologies (
>>> knowledge bases) supported by a common global
>>> schema as a "standard ontology for machines and
>>> people".
>>>
>>>
>>> Dear Azamat,
>>>
>>> A lot of people (including myself) believe that
>>> it is both extremely unlikely, not to mention
>>> undesirable, that a common global ontological
>>> schema will become a "standard ontology for
>>> machines and people". And indeed, if you look at
>>> the SW as it is today, you can already see
>>> thousands and thousands of alternative
>>> conceptualizations. This is why the NeOn project
>>> is developing a variety of practical solutions
>>> that focus on enabling the development,
>>> maintenance and use of networked ontologies,
>>> rather than assuming that some individual or
>>> organization will give us the 'global ontology'.
>>>
>>> Very Best Wishes
>>>
>>> Enrico Motta
>>>
>>> --
>>>
>>> The Open University is incorporated by Royal
>>> Charter (RC 000391), an exempt charity in England
>>> & Wales and a charity registered in Scotland (SC
>>> 038302).
>>>
>>>
>>>
>>
>>
>
>

Received on Wednesday, 7 January 2009 12:00:31 UTC

This archive was generated by hypermail 2.4.0 : Thursday, 24 March 2022 20:41:14 UTC