On 29/01/2011 12:39 PM, John F. Sowa wrote:
> On 1/29/2011 12:03 PM, Pavithra wrote:
>> "With Ontology say the same thing, mean the same thing, process the same
>> thing, everywhere"
>>
>> "Ontology enables semantic interoperability by presenting information
>> consistently across organizations and domains and machines"
> Both of those statements require a huge amount of qualification.
> As stated, they are false.
+1
> For many purposes, a highly underspecified definition is essential
> for interoperability. Names, addresses, and telephone numbers,
> for example, do not require a detailed specification of the
> nature of human beings, geography, or communication systems.
>
> Different applications may require radically different amounts
> of detail and formats for totally different purposes.
>
> For example, consider medical records, educational records,
> employment records, financial records, purchasing records,
> and Facebook sites for the same person. You definitely do
> *not* want the same kind of information specified in the
> same way in all of them.
>
You are also never going to get agreement between all of the software
vendors and industry standards groups about what describes a person. (01)
Too much of the discussion here seems to presume that we are going back
to the 1950s and 1960s were companies each built their own business
systems from scratch.
In the modern world, companies assemble a custom information structure
using software packages that come from several vendors.
If - BIG "if" - the world adopts ontology as the basis for future
generations of systems, then companies are going to have deal with many
ontologies that have some level of compatibility and some adapters that
allow concepts and information to flow from one to the other.
SAP will have a description of a person that is different from ORACLE's
HR view of the person which will be different from the payroll service's
view of the person or the LMS's view of the person or the insurance
company providing medical insurance or the government regulator that
monitors workplace safety and so on. (02)
However, these will all have to cooperate and provide adapters or
interfaces that allow the company's IT organization to make the whole
thing work.
Developers will need tools to "configure" ontologies to reflect the
company's view of the universe and to verify that these changes do not
affect interoperability. (03)
Is there a discussion about the metadata required to support the
management of such a lattice of ontologies?
This would seem to be a key thing to understand before trying to build a
useful repository and set of tools to use it. (04)
Ron (05)
> John
> (06)
>
> _________________________________________________________________
> Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
> Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
> Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
> Community Files: http://ontolog.cim3.net/file/work/OntologySummit2011/
> Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011
> Community Portal: http://ontolog.cim3.net/wiki/
> (07)
_________________________________________________________________
Msg Archives: http://ontolog.cim3.net/forum/ontology-summit/
Subscribe/Config: http://ontolog.cim3.net/mailman/listinfo/ontology-summit/
Unsubscribe: mailto:ontology-summit-leave@xxxxxxxxxxxxxxxx
Community Files: http://ontolog.cim3.net/file/work/OntologySummit2011/
Community Wiki: http://ontolog.cim3.net/cgi-bin/wiki.pl?OntologySummit2011
Community Portal: http://ontolog.cim3.net/wiki/ (08)
|