RE: [UCR] Coverage

Dave: I didn't catch what existing rule languages you were trying to interchange between.

My (probably wrong) impression is that you are proposing
RIF = new rule language + transport mechanism

Is not "transport" of rules is different from "interchange"?

For example: if you were using say SPARQL for your ontology mapping, what is the interchange issue in the use case? Surely you simply need to transfer the SRARQL rules with your ontology data between actors?

What am I missing here? Thanks.

Paul Vincent
Fair Isaac Blaze Advisor --- Business Rule Management
OMG Standards for Business Rules, PRR & BPMI
mobile: +44 (0)781 493 7229 ... office: +44 (0)20 7871 7229 
-----Original Message-----
From: Dave Reynolds [mailto:der@hplb.hpl.hp.com] 
Sent: Tuesday, February 21, 2006 12:32 PM
To: Vincent, Paul D
Cc: RIF
Subject: Re: [UCR] Coverage

The comment in the use case was mine so I recognize the use case is not 
clear cut.

However, I have two responses:

First, the use case is constructed so that rule exchange is actually 
needed. There are three actors, person with data in ontology O1, person 
wanting data in ontology O2 and a third party mediator who does the 
transform. These can be distinct actors [*] and it is a requirement that 
each actor may want to see and check how the ontology transform is being 
done, so the rules need to be at least visible to all (hence a role for RIF 
in publishing). In fact it is likely the rules will be created by one of 
the provider/consumer actors but executed by the third party mediator, 
again requiring rule exchange.

Whilst the use case is couched in terms of IT management services (because 
that's the specific area where it came up) it seems reasonable that on the 
semantic web there might be such general mediation services, i.e. services 
that take data from one or more semantic web sources, munge them and pass 
on the results with some value add. Rules would be one way that the 
mediators would either explain some of what they are doing to the data or 
be a means by which they might accept specifications for new "munging".

Second, this comment:

 > Which pretty much can apply to most rules!

is probably right. So what? Isn't part of the point of rule exchange so 
that one person can potentially reuse and learn from another person's rules 
if they are doing a similar job?

Last time I drew an analogy with SPARQL so let me pick XSLT this time. XSLT 
is used all over the place on the web to make XML flows work. If there are 
to be similar RDF data flows we need something like XSLT to help with all 
the impedance mismatches. A web rules language seems like a candidate for 
that. Whilst not the design centre for RIF it may be something RIF can help 
enable. We'll presumably know by the end of the use case process whether 
that's a reasonable ambition or not.

Dave

[*] In the application from which the use case is abstracted they are 
distinct actors though two of them were different parts of the same company.

Vincent, Paul D wrote:

> I have to concur with the commentary on the Message Transformation use case
> <<It is not clear to what extent this is a RIF use case as opposed to a rule use case, but it is no different in that respect from several of the currently proposed use cases.>>
> 
> Indeed: this is a rules use case, not an interchange one. Apart from the idea "if I need to transform an ontology, someone else might want to do the same transformation". Which pretty much can apply to most rules!
> 
> Paul Vincent
> Fair Isaac Blaze Advisor --- Business Rule Management
> OMG Standards for Business Rules, PRR & BPMI
> mobile: +44 (0)781 493 7229 ... office: +44 (0)20 7871 7229 
> 
> 
> -----Original Message-----
> From: public-rif-wg-request@w3.org [mailto:public-rif-wg-request@w3.org] On Behalf Of Ginsberg, Allen
> Sent: Monday, February 20, 2006 5:40 PM
> To: Dave Reynolds; RIF
> Subject: RE: [UCR] Coverage
> 
> 
> Hi Dave,
> 
> Thanks for your feedback.
> 
> Maybe it was a mistake to include Message Transformation under Third
> Party Rule-Interchange Services because the former actually does not
> involve interchange of rules so much as interchange of RDF-based data
> that has been massaged by rules. I think it was the fact that Message
> Transformation includes a third-party, namely the "Mediary Service,"
> that led me to see that connection.  
> 
> The rules in the Message Transformation use-case are RDF-transformation
> rules.  The RIF charter is very clear about the importance of
> compatiblity with RDF semantics.  So I am wondering if there shouldn't
> be a new general use case category with a title something like
> "Cross-Ontology RDF-Data Interchange."  This could be based on a
> fleshed-out version of your original use-case.  What we want is a
> detailed scenario. In your orignal use-case Wiki page you say: "A
> concrete narrative and example data set and rules could be provided if
> it becomes clear this is a useful enough case to expand in such
> detail,"  so I gather you could provide one.
> 
> Obviously this is something that the WG as a whole has to agree upon.
> 
> 
> Allen 
> 
> 
> -----Original Message-----
> From: public-rif-wg-request@w3.org
> [mailto:public-rif-wg-request@w3.org] On Behalf Of Dave Reynolds
> Sent: Monday, February 20, 2006 11:23 AM
> To: RIF
> Subject: [UCR] Coverage
> 
> 
> [Second email required to submit a "no" answer on the strawpoll, this
> time 
> for section "coverage".]
> 
> A primary use case for us is the use of rules to transform a set of RDF
> 
> statements from one ontology to another: 
> http://www.w3.org/2005/rules/wg/wiki/Message_Transformation
> 
> This use case has been grouped under general use case section 7 (Third 
> Party Rule-Interchange Services) but that section of the UCR draft
> document 
> seems specific to policy rules. One would not guess an ability of RIF
> to 
> express RDF transformations from the write up of that section.
> 
> The ontology transformation use case may be implicit as a special case
> of 
> Information Integration, if so it needs to more explicit in that
> section.
> 
> Dave
> 
> 
> 
> 
> 
>

Received on Tuesday, 21 February 2006 12:58:24 UTC