[Top] [All Lists]

Re: [ontolog-forum] blogic iswc keynote

To: "[ontolog-forum]" <ontolog-forum@xxxxxxxxxxxxxxxx>
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Fri, 18 Dec 2009 21:13:29 -0500
Message-id: <4B2C36C9.2070707@xxxxxxxxxxx>
Jim,    (01)

Some comments:    (02)

JR> SQL is basically a conjunctive normal form query language that can
 > be used to find individuals in a prescribed universe.    (03)

The WHERE-clause of an SQL query can express an arbitrary FOL statement,
which can be used in a query, a constraint, a trigger (a forward-
chaining rule), or a lambda expression that defines a virtual table.    (04)

The SELECT clause in front of the WHERE clause is syntactic sugar
for a lambda abstraction that identifies one or more fields as
formal parameters.  That lambda expression determines a relation
(AKA virtual table) whose values are to be extracted from the DB.    (05)

JR> [SQL] does not provide rules of inference.    (06)

That is true.  Any language of logic can be used with any rules
or assumptions that the users choose.  If they choose sound rules,
they get a monotonic logic.  If they choose nonmonotonic rules,
they get a system of plausible inference.  If they don't know what
they're doing, they can get in trouble.  But that's life.    (07)

JR> If a query returns no results, it is not possible to tell whether
> the query was inconsistent or there happened to be no individuals
 > in the database that satisfy the query.    (08)

That is true of any query system.  The following set is empty:    (09)

    {x | P(x) & ~P(x)}    (010)

JR> The "open world" semantics of the description logics make query
 > satisfaction quite a bit more difficult.    (011)

As with any language, the way a DL is used is up to the user.
As with SQL, any DL language can be used with either the open world
or the closed world assumption -- or any ad hoc combination.    (012)

For any logic notation, there are three independent issues:
the language, the rules of inference, and the model theory.    (013)

For different purposes, different choices can be made.  The tool
developers can provide support and guidance, or they can take
the attitude that they know better than any user and provide
a one-size-fits-all paternalistic solution.    (014)

John    (015)

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    (016)

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