ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Re Foundation ontology, CYC, and Mapping

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Fri, 12 Mar 2010 11:38:20 -0500
Message-id: <4B9A6DFC.1040903@xxxxxxxxxxx>
Sean,    (01)

Thanks for the citations.    (02)

I just wanted to make a few observations:    (03)

SB> Apologies for slow response to a couple of requests for sources
 > on semantic incompatibilities.  This is the table we generated
 > internally, based partly on older database work...
 >
 > [1] Aykut Firat, Information Integration Using Contextual Knowledge
 > and Ontology Merging. MIT (Sloan School of Management) Ph. D thesis,
 > September 2003.    (04)

For anyone who might like a 6-page article by the same author and
his thesis advisors:    (05)

    http://www.mit.edu/~bgrosof/paps/wits02.pdf
    Financial Information Integration In the Presence of Equational
    Ontological Conflicts    (06)

Some excerpts from that article and observations about them:    (07)

 > In this paper we first define what we mean by equational
 > ontological conflicts and then describe a new approach,
 > using Constraint Logic Programming and abductive reasoning...    (08)

1. First observation:  Note that their methods of representation
and reasoning are far richer and more expressive than anything
that could be done with just RDFS and OWL.    (09)

 > The likelihood of a single international accounting standard coming
 > to dominate anytime soon is quite slim. This is further complicated
 > by the complexities and localities involved in the accounting
 > practices of different countries (e.g. the UK views the proposed
 > standards as actually reducing the quality of their corporate
 > reporting.)    (010)

2. Don't expect that any proposed standard will magically solve all
problems -- it might even make some things worse.    (011)

 > ... when we collected Price Earnings Ratios for a specific company,
 > Daimler-Benz, from several financial sources on the same day the
 > numbers differed significantly, because of the differences in the
 > interpretation of earnings (see Figure 1.) A closer examination
 > reveals that these variations are not caused by erroneous reporting,
 > but attributable to definitional differences among data sources.    (012)

3. If they find such problems in data gathered about a single company
on a single day, just imagine the likelihood of deriving reliable
results from billions of triples defined by untold numbers of people
over an extended period of time.    (013)

 > ... ontological conflicts in accounting methods are quite widespread
 > not only between different countries, but also within the same
 > country... For example, the Wall Street Journal and S&P use different
 > methods to calculate the P/E Ratios for the Standard & Poor's
 > 500-stock index.    (014)

4. Attaching the URI of the definitional method might help resolve
 > this problem.  But note the following concern:    (015)

 > ... problems occur once companiesí financial numbers, crunched
 > by analysts, enter a vast information food chain, where they
 > are repeated, often without explanation, in hundreds of news
 > sources, and end up being used out of context.    (016)

5. As one example, companies deliberately sweep many qualifications
under the rug by using the term "Pro Forma earnings," which is
defined as "earnings before bad stuff."  Does anyone expect them
to highlight the bad stuff with a URI?    (017)

 > We have implemented ECOIN using the Eclipse Prolog Engine, its
 > extended CHR library, and the Java programming language.    (018)

6. Note that they are using Prolog to do their complex reasoning,
and that there is a version of Prolog available on the widely
used Eclipse platform for software development.    (019)

Prolog is an industrial-strength tool that is widely used by
government agencies and credit bureaus (Experian, for example).
Versions of Prolog are buried inside many software packages,
including Microsoft Windows.  A version of Prolog is the basis
for Wolfram's Mathematica system, which is used by Microsoft's
Bing search engine for processing complex queries.    (020)

At VivoMind, we translate RDF(S) and OWL to a high-speed version
of Prolog and run circles around the native tools for those
languages.    (021)

Bottom line: The tools and notations currently available for
ontology development are very useful as far as they go.  But
we are still at the beginning of where we need to go in the
development of semantic technology.    (022)

John Sowa    (023)


_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx    (024)

<Prev in Thread] Current Thread [Next in Thread>