Re: heterogeneous DBs

• Mail folder: Interlingua Mail
• Next message: James A. Fulton: "Re: heterogeneous DBs"
• Previous message: Bill Brayman: "heterogeneous DBs"
• In-reply-to: Bill Brayman: "heterogeneous DBs"
• Reply: James A. Fulton: "Re: heterogeneous DBs"
• Reply: Matthew L. Ginsberg: "Re: heterogeneous DBs"
• Reply: phayes@cs.uiuc.edu: "Re: heterogeneous DBs"

Date: Sun, 12 Sep 93 17:00:50 EDT
From: sowa <sowa@turing.pacss.binghamton.edu>
Message-id: <9309122100.AA02750@turing.pacss.binghamton.edu>
To: brayman@zuben.ca.boeing.com, interlingua@ISI.EDU
Subject: Re:  heterogeneous DBs
Cc: sowa@turing.pacss.binghamton.edu

There is a great deal of work going on in the database community to
introduce more semantics into DB systems.  In particular, the work
on ANSI standards using KIF has been initiated by the DB people,
not the KSE people.

I have been working with the ANSI X3H4 committee on IRDS (Information
Resource Dictionary Systems), which invited me to join their group
because they were interested in using conceptual graphs.  Since then,
I have proposed that we adopt both KIF and conceptual graphs as
exactly equivalent languages (identical semantic base and exact
mappings in both directions).

So Matt, you can blame me for dragging Fikes, Genesereth, Neches,
and Swartout to ANSI X3H4 meetings.  But don't worry, KIF and
CGs will not be a standard for AI.  As you can see from the recent
AAAI meetings, attendance is way down, the commercial companies
are pulling out of the exhibits, and the people with real applications
who need standards are staying away in droves.

So KIF and CGs will become a standard for commercial database and
CASE tools, but not for AI.  The whole AI field will dry up and
blow away in purity and innocence without any standards at all.

John Sowa