W3C home > Mailing lists > Public > www-rdf-rules@w3.org > September 2001

Re: Making Rules Out of RDF-Graphs (Re: What is an RDF Query?)

From: by way of <heflin@cse.lehigh.edu>
Date: Tue, 18 Sep 2001 13:31:44 -0400
Message-Id: <200109181731.NAA00789@tux.w3.org>
To: www-rdf-rules@w3.org
[freed from spam trap -rrs]

 Date: Tue, 18 Sep 2001 10:20:50 -0400 (EDT)
 From: Jeff Heflin <heflin@cse.lehigh.edu>
 Organization: Lehigh University
 Message-ID: <3BA757CA.4EEFB300@cse.lehigh.edu>
 To: Sandro Hawke <sandro@w3.org>
 CC: "Peter F. Patel-Schneider" <pfps@research.bell-labs.com>,

Sandro Hawke wrote:
> Maybe I'm just a guy holding a hammer called "Horn" but it seems
> useful to me.  There are a variety of ways one should evaluate a
> proposed technical standard, and this seems like a useful angle,
> somewhere between a pure-theory analysis and running code.
>      -- sandro


I agree that Horn logic is very useful. The SHOE Language [1] (which was
developed by Jim Hendler, Sean Luke, and myself) was based on Horn
logic. This made it easy to use XSB (or other deductive database
systems) as complete reasoners for SHOE. I think the advantage of
deductive database systems is that, unlike many AI systems, they are
designed to reason over large detests. That being said, DAML+OIL is
based on Description Logics (DLs), which can express some things that
cannot be said in Horn logic (similarly, Horn logic can express some
things that DLs can't). A deductive database will probably perform
better than a DL system, but if the DL provides additional expressivity
that users need, then the extra speed is not very useful. Therefore, I
believe it is reasonable to provide two languages: a Horn-based rule
language, and the DL-based DAML+OIL. Then users can choose the language
most appropriate for their needs. However, we may want to consider
creating a language that is the intersection of these two languages
(such a language would be more expressive than RDFS), so that they have
a common foundation (above and beyond RDFS). Alternatively, we could
change RDFS to include all features that are common to the two logic

Jeff Heflin
Assistant Professor
Lehigh University

[1] http://www.cs.umd.edu/projects/plus/SHOE/
Received on Tuesday, 18 September 2001 13:31:51 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 2 March 2016 11:10:13 UTC