Let me clarify: (01)
BG:
If we talk to the developers who know their own systems
inside-and-out, and allow them to incorporate ontologies via tools... (02)
JS:
Right. But no ontology today will help anybody "build and maintain an
application quickly." (03)
My apologies if I didn't explain myself. I'm not proposing an ontology
to develop tools, but instead tools that allow developers to work WITH
ontologies. Speaking their language essentially. The reason I brought
up open source is because I see why developers and their companies
have been adopting new open source technologies successfully. Years
back MVC was only know in the software engineering context, now it's
known by every developer (give or take). (04)
The more developers adopt them, the more use-cases they produce.
Companies have been seeing this trend and open sourcing parts of their
systems, or opening up API's to their data. This only makes those
technologies (and their owners) richer. To take advantage, developers
have been learning more and more about data-mining and other ways to
manage data. My background is in machine-learning so I've seen how
novice data users approach data-mining. I'd like to see this happening
for ontologies. (05)
Open Data is an opportunity to promote semantics as well. The OpenData
repositories, free and pay-per-use, are growing. I'd love for them to
start supporting the types of semantic tools Bioportal and OOR
provide. If they provide data via API's that support semantics and
allow incorporating them quickly, companies will start seeing the
benefit of viewing them through this "semantic view". (06)
So two ways to help developer adopt ontologies are: (07)
1) Develop/promote the tools that allow them to work with semantics in
one shape or another. Whether it's OWL, RDF, etc, these can be the
steppingstones to the acceptance of ontologies like SUMO and Cyc. (08)
2) Give them only the parts of ontologies they need and can work with
right away. Produce subsets of ontologies to solve a particular
problem in manageable chunks. As simple as they are, FOAF and
GoodRelations are great examples, and relate to topics their adopters
understand, personal relationships between people and companies. Let's
promote the technologies people are using to incorporate these
ontologies. (09)
On 4 March 2011 15:10, John F. Sowa <sowa@xxxxxxxxxxx> wrote:
> Bart,
>
> The main point I was trying to make is that you can't measure
> the result of an ontology by itself.
>
> For example, you could take an old mainframe application that
> drives a "green screen", write a new front end, and get a web
> application. But you can't use an ontology to upgrade any kind
> of running application. You have to start from scratch without
> any guarantee that the result will do anything useful.
>
>> It's not about promoting specific ontologies, or even ontologies
>> in general. It's about promoting the tools to use with ontologies
>> first, and for this audience specifically promoting Cyc or SUMO
>> is what creates the blank stares in the first place.
>
> Historical interlude: Back in the 1990s, Cyc was supported by
> some large corporations and gov't agencies. They all employed
> people who had PhDs in comp. sci., even with specialties in AI.
> There were no blank stares, because they had already bought
> Doug Lenat's sales pitch.
>
> Every one of those groups had a complete set of Cyc software and
> the full ontology. I talked with some of the people at those
> companies and agencies. But none of their employees were able
> to develop a single deployed application. One of them said
> "Everybody who spent any significant time working with Cyc
> was let go, and I don't believe that's a coincidence."
>
> I admit that Cyc has improved quite a bit during the past decade.
> But one of the people at a location where Cyc is used said that
> there is a "huge disconnect" between the skills required to work
> with Cyc and the skills required for any other software.
>
>> Ease of development is what the Ruby on Rails framework provides.
>> It's very popular only because it allows developers to build and
>> maintain an application quickly.
>
> Right. But no ontology today will help anybody "build and maintain
> an application quickly."
>
>
> What tools? Where is the ontology equivalent of "Ruby on Rails"?
>
> John
>
> _________________________________________________________________
> 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/
> (010)
--
Bart Gajderowicz, MSc.
Ryerson University
http://www.scs.ryerson.ca/~bgajdero (011)
_________________________________________________________________
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/ (012)
|