[iaoa-education] proposal: new agenda item for today == FW: [ontolog-forum] Proposed ISO standard for ontology

Frank Loebe frank.loebe at informatik.uni-leipzig.de
Thu Jul 6 13:24:58 CEST 2017 

Dear all,

as we have a meeting today, I think we should discuss the current 
discussion on the term list at [ontolog-forum] [1] (see below) and a 
reaction to that, on the list and in the wiki, e.g. by a disclaimer.

Best regards,
Frank

PS: offline for some hours


[1] https://groups.google.com/d/msg/ontolog-forum/mK7s90BjHDM/rDjgGsT6BgAJ



------ Original Message ------
From:    David Price <dprice at topquadrant.com>
Sent:    2017.07.06 10:40 +0200
To:      ontolog-forum <ontolog-forum at googlegroups.com>
Subject: Re: [ontolog-forum] Proposed ISO standard for ontology

> FWIW I agree 100% with Pat wrt the terms list. I had a quick look at a few there are mostly really terrible (i.e. they would make the current situation worse).
> 
> Look at the definition of “Ontology” for a prime example that will do nothing but add confusion:
> 
> Ontology:  An ONTOLOGY is a representational artifact, comprising a taxonomy as proper part, whose representational units are intended to designate some combination of universals, defined classes, and certain relations between them
> 
> No consensus outside BFO would ever be reached on that (i.e. none of that text would survive review by anyone outside the BFO community).
> 
> Best stop that exercise before any more time is wasted, and rather spend time on other fruitful activities.
> 
> If for some reason people choose to continue, then every noun must have an adjective to specify context … so “BFO Ontology”, not “Ontology”. No point in saying “Ontology” and then list 10 completely conflicting definitions … that adds nothing wrt wider consensus or clarity.  That must follow through into the definitions too. For example, the "BFO Ontology" definition depends on very specific definitions of the terms it uses and I’m sure other of less-than-useful definitions would also appear for those terms in the list. How would anyone not deeply involved even start to parse such definitions without a clear context adjective used everywhere, all the time?
> 
> I’m not 100% sure of the intended audience, but I would certainly *never* point an enterprise customer I was trying to convince of the value of a semantic approach to such a term list (even specified as I suggested) as it would provide nothing but ammunition for the naysayers.
> 
> Cheers,
> David
> 
> UK +44 7788 561308
> US +1 336 283 0606
> 
> 
> 
> 
>> On 5 Jul 2017, at 22:32, Todd Schneider <tjschneider at covad.net> wrote:
>>
>> Pat,
>>
>> I can understand you comment about 'consensus on terminology',
>> but in the context of the IAOA term list and its intent, I think another
>> look may be in order (i.e., I may have failed to properly explain myself).
>>
>> The original intent of the list of terms was to collect those (natural language)
>> terms deemed useful to understanding some of the background of the various
>> disciplines that contribute to ontology and that appear in many source materials
>> (relating to ontology) and their more common definitions.
>>
>> The assumption was that many terms do have different definitions and being able
>> to provide these variations (and something of their context) would be helpful in
>> promulgating a better understanding of ontology and its underpinnings.
>>
>> At present the IAOA term list is undergoing revision to weed out less relevant
>> entries (e.g., BFO, Common Logic, OWL, etc.).
>>
>> The consensus I was (ambiguously) referring to was on the terms themselves
>> and possibly the various definitions that should be included. Not consensus
>> on a single definition (per term).
>>
>> Todd
>>
>> P.S. The best uses I've found from foundational ontologies (aka Upper ontologies)
>>        are the explanations, motivations, and reasoning behind the various decisions
>>        made in their creation.
>>
>>        In application, those explanations have been more flexible and interoperable
>>        the ontologies themselves.
>>
>>
>> On 7/5/17 4:33 PM, Pat Hayes wrote:
>>>> On Jul 5, 2017, at 7:38 AM, Todd Schneider <tjschneider at covad.net> wrote:
>>>>
>>>> John,
>>>>
>>>> What I hope to gain from the IAOA meeting is better understanding
>>>> on what a 'theory of ontology' (in the context of information systems)
>>>> should encompass and what that may entail and possibly extract better
>>>> design principles.
>>> That sounds eminently reasonable and a laudable goal. But…
>>>
>>>>  From a practical perspective l'll be pushing for consensus
>>>> on common terminology (i.e., see the IAOA terminology list
>>>> http://iaoaedu.cs.uct.ac.za/pmwiki.php?n=IAOAEdu.TermList).
>>> …this is a disaster. Both the selection of terminology to standardize and the definitions offered are completely idiosyncratic, in some cases to the point of seeming wilfully obscure, and are completely influenced and dominated by the OBO philosophical tradition, ignoring all other uses of words like “class". To standardize on these would be simply to create another useless silo which will be ignored by some and enthusiastically treated as a gospel by others.
>>>
>>> Pat Hayes
>>>
>>>> Also, the discussions should be entertaining.
>>>>
>>>> Todd
>>>>
>>>>
>>>> On 7/4/17 11:35 PM, John F Sowa wrote:
>>>>> Dear Matthew and Todd,
>>>>>
>>>>>>> [JFS] My main criticism of the current draft is that it says
>>>>>>> nothing about the complex issues of knowledge sharing and semantic
>>>>>>> interoperability that have been analyzed and debated for the past
>>>>>>> half century.
>>>>>> [MW>] Indeed it does not, for the simple reason that is not the
>>>>>> problem it is trying to solve. It is trying to do something much
>>>>>> smaller, which is to set some standards for Top Level Ontologies
>>>>>> that hopefully ontology developers will aspire to meet.
>>>>> Unfortunately, that goal ignores the purpose of an ontology:
>>>>> support interoperability among independently developed systems
>>>>> -- *especially* the trillions of dollars of software that has
>>>>> no explicit ontology of any kind.
>>>>>
>>>>> This goal raises fundamental questions about how to design a TLO
>>>>> that has that level of flexibility.  There are several options:
>>>>>
>>>>> 1. Edict a single TLO, which serves as the hub of a family of
>>>>>     spokes.  Each TLO is the hub of a silo that is incompatible
>>>>>     with every competing silo -- and with the multi-trillions
>>>>>     of dollars of legacy software.
>>>>>
>>>>> 2. Develop a theory for designing ontologies in a systematic way
>>>>>     that facilitates interoperability.  This is the microtheory
>>>>>     hypothesis of CYC and most large modern ontologies.
>>>>>
>>>>> 3. Ignore the top levels and focus on interoperability at the
>>>>>     mid levels and lower levels.  This is the basis for using
>>>>>     WordNet and other lexical resources to align the data at the
>>>>>     level of ordinary language.  This method has been moderately
>>>>>     successful for legacy systems from the punched-era of the
>>>>>     1890s to nearly every system connected to the WWW today.
>>>>>
>>>>> 4. Develop mathematical methods for finding relationships among
>>>>>     independently developed systems.  Examples include DOL and
>>>>>     other kinds of mathematical techniques.  This method could
>>>>>     help #2 and #3 above.  It might even relate some silos in #1.
>>>>>
>>>>> Option #1 (the hub & spoke model) is a strategy to promote
>>>>> incompatible silos.  I believe that this strategy is the primary
>>>>> reason why practical developers ignore ontology.
>>>>>
>>>>>>> [JFS] Design Part 2 as a collection of microtheories from various
>>>>>>> sources.  The BFO top-level is small enough to be a microtheory.
>>>>>>> Other microtheories may be added to Part 1 from any source.
>>>>>> [MW>] Not part of the scope. The parts after Part 1 will be other
>>>>>> TLOs that wish to be standardised as alternatives to BFO. Alternatively,
>>>>>> for existing standard TLOs, they can claim conformance to the standard.
>>>>> In short, the scope (copy below) implies that the standard will
>>>>> consist of multiple competing TLOs, each of which is designed to
>>>>> be the hub of a silo that is incompatible with every other silo.
>>>>>
>>>>> Todd
>>>>>> Will you be attending the IAOA Summer Institute on Upper Ontologies,
>>>>>> to be held in Toronto, this August?
>>>>> No.  Just the thought of listening to people claim that their silo
>>>>> is better than anybody else's silo provokes retchophobia.  I have
>>>>> better things to do.
>>>>>
>>>>> John
>>>>> __________________________________________________________________
>>>>>
>>>>> An excerpt from the scope Part I:  (Comment: silos by design.)
>>>>>
>>>>> This International Standard focuses on ontologies to be used as
>>>>> resources designed to support the interchange of information among
>>>>> heterogeneous computer systems. It specifies a hub-and-spokes
>>>>> architecture for ontology development and sets forth the requirements
>>>>> an ontology shall satisfy if it is serve as hub in such an architecture.
>>>>>
>>>>> The following are within the scope of this International Standard:
>>>>>
>>>>> • Specification of how ontologies used for data retrieval, integration
>>>>> and analysis can be combined into modular suites of mutually consistent
>>>>> and non-redundant ontologies.
>>>>>
>>>>> • Specification of the hub-and-spokes structure of such ontology suites.
>>>>>
>>>>> • Specification of the role of definitions in a hub-and-spokes ontology
>>>>> architecture.
>>>>>
>>>>> • Specification of the requirements for an ontology to serve as hub in
>>>>> such a structure.
>>>>>
>>>> -- 
>>>> All contributions to this forum by its members are made under an open content license, open publication license, open source or free software license. Unless otherwise specified, all Ontolog Forum content shall be subject to the Creative Commons CC-BY-SA 4.0 License or its successors.
>>>> --- You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-forum+unsubscribe at googlegroups.com.
>>>> For more options, visit https://groups.google.com/d/optout.
>>>>
>>>
>>
>> -- 
>> All contributions to this forum by its members are made under an open content license, open publication license, open source or free software license. Unless otherwise specified, all Ontolog Forum content shall be subject to the Creative Commons CC-BY-SA 4.0 License or its successors.
>> --- You received this message because you are subscribed to the Google Groups "ontolog-forum" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to ontolog-forum+unsubscribe at googlegroups.com.
>> For more options, visit https://groups.google.com/d/optout.
>

More information about the iaoa-education mailing list