Re: RDF query and Rules - my two cents

Hi Jim et al,

At 11:55 AM 11/11/2003 -0500, Jim Hendler wrote:

>Dan et al-
>  Been thinking hard about this - here's my two cents:
>
>1 - I think there is a clear and present need in the RDF community for a 
>way to essentially request a set of triples from a remote store -- 
>essentially an RDF remote access API.   We are already seeing many cases 
>(including but not limited to RSS feeds) where the sharing is very 
>powerful, and if we could do that more geenrally, it would help many 
>projects.   As more RDF-based web portals grow, this ability becomes more 
>important -- for example, I have some US computer scientists described on 
>my web site in OWL, the AKT project in the UK has many British computer 
>scientists described in OWL.  We cannot, hwoever, get information from 
>each other's backend stores without negotiating our own protocols - and 
>this makes it hard to get N-way agreement.
>  So this is an aspect of query which I will call "Remote Access" and I 
> think we are ready for a WG on this.  Note that this would probably not 
> be very exciting for logicians, DL fans, etc. because I would expect this 
> to be simply a triples-exchange-mechanism over HTTP, not a real "query" 
> langauge
>
>2 - I think there is the possibility of creating a query group based on 
>your charter, it would still need the above to exist to work, but would 
>add some logical notions and possibly blur with rules.   I think that 
>holding a workshop or two to try to tease apart these issues is needed - 
>as evidenced by the discussion on this mailing list.
>
>3 - with due respect to the folks involved, I think the current Rules 
>charter is way too broad,

I agree. It was written rather in a rush by the W3C folks. It clearly needs 
work.

>and a WG started in that area would thrash for a long time.

That probability is fairly high if you assume the current charter draft is 
a mature document providing a close indication of doability, technical 
maturity of the area, and a best possible agenda.  But that is not the case 
-- it's only a still-pretty-raw first draft.

>I think we need activity to determine how to limit such a charter to 
>something doable,

Yes.  But I think we don't *need* to have a workshop and wait an *extra* 6+ 
months *before* getting going on having a Working Group.
(Although a workshop would probably be helpful as part of kicking off a 
committed Working Group effort.)
What we do need are some revisions to the charter of the Working Group.  I 
think there's a very good chance consensus can be reached on that in much 
less than an extra 6 months, with confidence in doability.

>or find a more viable "de facto" standard to build from -- as far as I can 
>see we dn't have a consensus in that area --

On the contrary, I think we do have exactly such a consensus starting point 
as a first step:
the Joint Committee's rules proposal, which is basically to layer 
restricted Horn rules cf. RuleML on top of OWL.  This is
currently in late draft and within the next few weeks is planned to be 
submitted as a W3C Note.
It represents a consensus built up over the last 15+ months in the Joint 
Committee in close cooperation with the RuleML Initiative
and the DAML program, as well as in looser cooperation with the Semantic 
Web Services Initiative over the last 11 months.
It reflects directly and indirectly inputs from dozens-to-hundreds of 
researchers/developers at several-dozen companies/institutions.

Benjamin


>I would hope EU/DARPA/NSF or others might host a forum in which such a 
>consensus could emerge - otherwise I would think the group would be aiming 
>for something so ambitious that it might miss the short-term targets, 
>without benefit of enough maturation to meet the long-term needs.
>So I would propose:
>
>i. W3C charter a "RDF Web Remote Access" WG in the near future - limited 
>charter, short time-frame, get an API to rec.
>ii. W3C sponsor a workshop on "Rules and Queries" in early 01 to explore 
>the range of issues in these charters.  If there was consensus that a 
>single WG could do these together then that would be a good possibility 
>for a Wg to start in Fall.
>
>Note that RDF Core and OWL will need to stay together for about 6 months 
>after PR to do debugging and maintenance, so that starting too many WGs 
>immediately is likely to cause a real strain on the limited number of 
>people we have.
>
>so - Remote access now; Rules/Query workshops and some start in that space 
>in the Fall, would be my proposal...
>  -JH
>
>--
>Professor James Hendler                   http://www.cs.umd.edu/users/hendler
>Director, Semantic Web and Agent Technologies     301-405-2696
>Maryland Information and Network Dynamics Lab.    301-405-6707 (Fax)
>Univ of Maryland, College Park, MD 20742          240-277-3388 (Cell)

________________________________________________________________________________________________
Prof. Benjamin Grosof
Web Technologies for E-Commerce, Business Policies, E-Contracting, Rules, 
XML, Agents, Semantic Web Services
MIT Sloan School of Management, Information Technology group
http://ebusiness.mit.edu/bgrosof or http://www.mit.edu/~bgrosof

Received on Tuesday, 11 November 2003 15:06:24 UTC