On 3/4/2011 10:26 PM, Pat Hayes wrote:
> OWL is based on description logics, which have a much wider usage
> in real-world applications than all of AI KR work put together. (01)
No DLs are integrated with the mainstream software development tools.
They are all niche tools. (02)
Among DLs, PowerLoom was very widely used, but Bob MacGregor, the
chief designer, was drummed out of the DL community as a heretic.
He dared to say the obvious: "No users ever ask for decidability
-- but they all want more expressive power." (03)
Rule-based languages are very widely used for mission-critical
applications. One example is Experian, which uses Prolog to
check everybody's credit rating. They use Prolog so much that
they bought Prologia, the company founded by Alain Colmerauer,
who invented Prolog. (04)
Unfortunately, the only approved reasoning method available for
the SW is OWL. There is a box labeled RIF, but no tools that
support it. SWRL attracted a fair amount of attention, but
the decidability thought police squashed it because it was
(Horror of Horrors!) undecidable. (05)
> LISP is hardly used by anyone. Much as we all regret this,
> it is the truth. (06)
I agree. But it is inexcusable that the AI developers caved in
to people who had a horribly misguided hypothesis about how to
design a language. (07)
> Twice in the last year I have had young programmers telling me
> excitedly about this really neat new system they have discovered
> called LISP, and have I heard of it? History seems doomed to
> repeat itself over and over again. (08)
Of course. It's obvious to anybody who has any taste. (09)
> RDF was developed 1999-2004, JSON in 2006... (010)
JSON is JavaScript Object Notation, which was available on every
browser. HTML also supported the <script> tag, which could be
(and still can be) used to bracket any decent notation. (011)
> The RDF revision now under way has JSON integration as
> a top priority and one of the first three action items. (012)
Tim Bray, who collaborated with Guha in designing RDF, admitted
that he made a mistake on his web site in 1999: (013)
http://www.tbray.org/ongoing/When/200x/2003/05/21/RDFNet (014)
It is inexcusable that the W3C did nothing about that mistake
for over a decade. We should be grateful to Google for rejecting
RDF, but it's a shame that so many AI people caved in. (015)
> Meanwhile, deplore it as we may, XML syntax is more widely used
> than all other notations ever developed. (016)
XML is an excellent notation for tagging documents. That was
the purpose of GML (which I used at IBM since the 1970s),
SGML, and HTML. And that is still its "sweet spot". (017)
> UIMA isn't a Krep notation, it is a system model for applications
> designed to process unstructured text. (018)
That is true, but it uses XML to express information in a more
rational way than RDF: (019)
<author>William Shakespeare</author> (020)
which isn't too far from LISP in efficiency: (021)
(author "William Shakespeare") (022)
John (023)
_________________________________________________________________
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/ (024)
|