Hi Ed --
Thanks for your overview of the state of play in "rules".
You wrote...
As I said earlier, "rules" is just too big an umbrella. Validating data
sets, transforming data, directing action, performing directed
reasoning, providing guidance, and capturing and interpreting policy and
regulation are all applications of "rules" that involve different but
overlapping technologies. When you know what kind of problem you want
to solve, then you can talk about the "rules" concepts and activities
that are relevant to solving that class of problem. That is why the
RuleML effort resulted in 5 languages. And for the same reason, we
probably won't see "widely adopted standards", because no single "rules
technology" addresses more than a small part of the spectrum of "rules
applications".
You have stated very clearly a problem that has been worrying me (and maybe others) since the early days of RIF.
Back then I proposed a way of unifying diverse rules systems on the web:
www.w3.org/2004/12/rules-ws/paper/19
The basic idea is to get the various commercial and other rule systems
talking to each other at the input-output level, and to map out how to
gradually make their black boxes more scrutable. The proposal
also tried to address the problem of the conceptual complexity
and auditability of interacting rules systems on the Web, by bolting
English commentary containing variables to the input-output messages.
Maybe the proposal is worth revisiting? What do you and the list folks think?
Thanks, -- Adrian
Internet Business Logic
A Wiki for Executable Open Vocabulary English
Online at www.reengineeringllc.com Shared use is free
Adrian Walker
Reengineering
On 10/18/07, Ed Barkmeyer <edbark@xxxxxxxx> wrote:
Jyoti,
you wrote:
> Is there a study and/or a general opinion about which of these 3
> activities is more widely adopted, practical (in terms of
> applications) etc.?
Apart from propaganda, not that I know of. The fact is that they are
all essentially "academic" activities at this point. I don't think
there is any commercial development of SWRL, and there was no commercial
involvement in RuleML until the last year or so, and even that is more
"academic spinoff" and government-sponsored work. Obviously the W3C
effort (RIF) has the most caché, and it has both academic and commercial
participation, but it isn't yet stable enough to have implementations.
SWRL was published as a W3C member submission in 2004, has not gained
any greater status, and the only implementations I have heard of were
academic prototypes (someone's thesis). I know of only a few tools that
speak some dialect of RuleML.
But OWL and RDF were government-sponsored efforts with almost entirely
academic participation until they achieved visibility as W3C standards,
and then the floodgates opened. So "widely adopted" is really a matter
of filling a need and achieving buzzword status, and my money would be
on RIF, but that it is the least advanced work of the lot.
There are two other activities I omitted, both recently published and
available online from the Object Management Group (
www.omg.org):
- the Production Rules Representation specification is a simple model of
condition/action rules that has a standard XML representation per OMG
XMI. It was developed by major commercial rules engine vendors -- ILOG,
Fair-Isaac, Computer Associates, et al. Their blurb says they intend to
use a subset of RIF as the official XML exchange form when RIF is done.
- the Semantics of Business Vocabulary and Rules specification is
purportedly about the capture and exchange of rules, but its formal
basis is foggy. What it seems really to be about is getting a formal
form in which "what business people said" can be captured and exchanged.
So it is all about linking the terms used in rules to formal and
natural language definitions. In that way, it is much closer to the
Attempto "controlled English" kind of thing. There was no clear intent
that the formalized statements should be inputs to any class of
reasoning engine, although at least two of the participants have some
engine that will be able to do something meaningful with some subset of it.
The PRR will almost certainly have implementations by several commercial
rules engine vendors, so that they can sell diverse decision support,
workflow management, EAI and software generation applications to the
same major customers. (That is why they got together to make the
standard.) As to SBVR, IMO, the "vocabulary" part is more likely to be
valuable than the "rules" part.
As I said earlier, "rules" is just too big an umbrella. Validating data
sets, transforming data, directing action, performing directed
reasoning, providing guidance, and capturing and interpreting policy and
regulation are all applications of "rules" that involve different but
overlapping technologies. When you know what kind of problem you want
to solve, then you can talk about the "rules" concepts and activities
that are relevant to solving that class of problem. That is why the
RuleML effort resulted in 5 languages. And for the same reason, we
probably won't see "widely adopted standards", because no single "rules
technology" addresses more than a small part of the spectrum of "rules
applications".
-Ed
--
Edward
J.
Barkmeyer Email:
edbark@xxxxxxxx
National Institute of Standards & Technology
Manufacturing Systems Integration Division
100
Bureau Drive, Stop
8263 Tel:
+1 301-975-3528
Gaithersburg, MD
20899-8263 FAX:
+1 301-975-4694
"The opinions expressed above do not reflect consensus of NIST,
and have not been reviewed by any Government authority."
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Subscribe/Config: 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 Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (01)
|