W3C home > Mailing lists > Public > public-rif-wg@w3.org > March 2010

Re: [RIF UCR] Publication plan for the next UCR version

From: Leora Morgenstern <leora@cs.nyu.edu>
Date: Tue, 16 Mar 2010 09:57:24 -0400
Message-ID: <417caf6b115facfdcae7de02bc27b3b1.squirrel@webmail.cims.nyu.edu>
To: "Adrian Paschke" <adrian.paschke@gmx.de>
Cc: "'Leora Morgenstern'" <leora@cs.nyu.edu>, "'Chris Welty'" <cawelty@gmail.com>, "'Public-Rif-Wg'" <public-rif-wg@w3.org>
Hi Adrian,

Your email brings up several interesting points.

With regard to the purpose of the use cases: I  do realize how they came
about, historically. (I was one of the developers of the original use case
on medical decision making.) You are right that they were intended to
motivate the development of RIF. Along the way, they also seem to have
acquired other uses, including motivating requirements. At some point,
moreover, uses cases often function as checkpoints of sorts: being able to
express a use case within whatever language/system has been developed
lends validity to the enterprise. Much the same happened here.

This is why there has been for a while an effort to produce code for the
use cases. Not having code would seem to imply that the use case
*couldn't* be handled within the current RIF dialects. But there seem to
be use cases for which there is currently no code, which ought to be
formalizable within RIF. Examples include: Interchanging Rule Extensions
to OWL,  BPEL Orchestration of Rule-Based Web Services (it's currently in
some sort of Prolog-like formalization, which isn't legal BLD), 
Integrating Production Systems in Cases of Business Mergers, Managing
Inter-Organizational Business Policies and Practices. There are also cases
like Access to Business Rules or Supply Chain Partners, which I believe
can be done in FLD (though I have to check the latest version of FLD to
make sure).

So, I do think that it would be worthwhile to modify the UCR to formalize
these cases, which I hope to do in the next week.

With regard to the categorization: You are correct that there are some
cases which can be formalized in multiple dialects. There are several ways
we could handle this. (1) We could add a paragraph specifying that some
use cases can be formalized in several dialects, and then mention that
fact when the particular use cases are given. (2) We could do away with
the categorization, but mention those use cases that can be formalized in
multiple dialects.

I believe  (1) is preferable as a pedagogical device: It helps readers
more easily see how the different dialects can be used.

This segues into the issue of having a tutorial/primer document for RIF,
and what the possible connections would be between the UCR document and
the tutorial/primer document. I'll address this in the next email.

Best regards,
Leora


On Tue, March 9, 2010 11:27 am, Adrian Paschke wrote:
> Hi Leora,
>
> Since you started editing on UCR I have added you to the editors list of
> the
> document:
>
> http://www.w3.org/2005/rules/wiki/UCR
>
> I have looked at your restructuring of UCR which now lists the use cases
> for
> the three existing dialects BLD, PRD and FLD. I'm not sure if that is what
> we initially intended UCR for.
> The use cases which we selected from the 50 members submissions and which
> we
> further detailed in the UCR document where intended to demonstrate the
> need
> and usability of a W3C RIF standard in general.
>
> For those use cases which can be already formalized with the existing RIF
> dialects I had added code examples and respective buttons to show/hide
> them,
> in order to "to provide a reference to the design of RIF and a guide for
> users and implementers to the current technical specifications of RIF
> dialects." (see http://www.w3.org/2005/rules/wiki/UCR#Introduction)
>
> I remember that we discussed in the working group that we remove all code
> examples from use cases which currently cannot be represented with the
> existing RIF dialects, e.g. because they require some expressive logical
> formalism such as event calculus. However we said we want to keep these
> use
> cases for the following reason (see
> http://www.w3.org/2005/rules/wiki/UCR#Introduction):
>
> "RIF-UCR also delivers a structured context for formulating future
> technical
> specifications of further RIF dialects. Each dialect targets at a cluster
> of
> similar rule languages and enables platform-independent interoperation
> between them (via interchange of RIF rules). The presented use cases
> illustrate some of the principal ways in which RIF can provide benefits."
>
> I think this second point of reference is important for RIF as an
> interchange format in general.
>
>
> Another problem with the restructuring into BLD, PRD, and FLD use cases is
> that some of them can be represented in both BLD and PRD. I remember we
> had
> long discussion about the business rule use cases which can be represented
> in both BLD and PRD, and the presentation syntax for rules using ":-" "->"
> or "<--". The new structure might suggest that you cannot represent it in
> the other dialect. For instance, use case 4.1. can be represented in BLD
> and
> PRD:
>
> http://www.w3.org/2005/rules/wiki/UCR#Negotiating_eBusiness_Contracts_Across
> _Rule_Platforms
>
> So I think we should change it back to the original structure and just
> update the code examples to the latest syntax from the latest dialect
> specification. For those use cases which cannot be represented yet with
> the
> existing RIF dialects we can add a note. Since new dialects are on their
> way, such as a logic dialect with negation, some of them can be
> represented,
> soon.
>
> Best,
>
> Adrian
>
>
> -----Ursprüngliche Nachricht-----
> Von: public-rif-wg-request@w3.org [mailto:public-rif-wg-request@w3.org] Im
> Auftrag von Leora Morgenstern
> Gesendet: Dienstag, 2. März 2010 14:42
> An: Chris Welty
> Cc: Public-Rif-Wg
> Betreff: Re: Next few weeks in RIF
>
> Chris,
>
> Thanks for the reminder. I did some re-organizing of the document in the
> summer and fall, dividing the use cases into BLD, PRD, and FLD. In some
> cases, the division was mandatory (e.g., there is at least one case that
> must be done in FLD because it makes explicit reference to the concept of
> belief); but there were some cases that could go either way (e.g., some
> cases seem more naturally expressed in PRD although they could probably be
> represented in BLD.)
>
> I also put in a bunch of editors' notes as a reminder to myself of work
> that I still need to do, specifically translating cases into the required
> dialect, making notation consistent, etc.
>
> I am booked solid this week and traveling next week, but can get to this
> the week of March 14th, and would aim to finish within a week from that
> time.
>
> Best regards,
> Leora
>
> On Mon, March 1, 2010 2:51 pm, Chris Welty wrote:
>>
>> RIFWG,
>>
>> We'd like to make a push for the next few weeks to get ready for
>> transition.  We have some decisions to make, some work to do on the
>> documents, etc.
>>
>> So, we will resume weekly telecons for about a month, starting tomorrow.
>>
>> What is the status of the "other" documents, should we change or update
>> them?  Are they in their final form?
>>
>> UCR: Leora?
>> XML-data: Christian?
>> OWL2/RL: Dave?
>> Overview: Michael?
>> Test: Stella?
>>
>> -CC&S
>>
>> --
>> Dr. Christopher A. Welty                    IBM Watson Research Center
>> +1.914.784.7055                             19 Skyline Dr.
>> cawelty@gmail.com                           Hawthorne, NY 10532
>> http://www.research.ibm.com/people/w/welty
>>
>>
>
>
> --
> Leora Morgenstern, Ph.D.
> http://www-formal.stanford.edu/leora

>
>
>


-- 
Received on Tuesday, 16 March 2010 13:58:11 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 16 March 2010 13:58:13 GMT