Semantic Web Interest Group IRC Chat Logs for 2004-03-02

00:41:51 <tav> tav is now known as tav2

00:42:01 <tav2> tav2 is now known as tav

01:49:40 * sh1m|cannes yawns

03:23:59 <Davey> Davey is now known as D[a]vey

05:59:49 <SirJordie> SirJordie is now known as SirKatz

07:05:37 <CaptSolo> morning!

08:01:57 <dajobe> hello

08:02:12 <dajobe> the SW IG room is filling up slowly

08:02:34 <CaptSolo> all the beer drinkers in Cannas are waking up

08:02:38 <CaptSolo> greetings :)

08:02:50 <dajobe> the TAG started meeting at 8am local

08:02:52 <CaptSolo> s/Cannas/Cannes

08:04:01 <dajobe>http://www.w3.org/2001/sw/interest/meetings/tp2004.html

08:04:03 <dc_rdfig> A: http://www.w3.org/2001/sw/interest/meetings/tp2004.html from dajobe

08:04:15 <dajobe> A:|1-2 March 2004 Meeting of the W3C Semantic Web Interest Group

08:04:16 <dc_rdfig> Titled item A.

08:04:28 <dajobe> logger, chump A

08:04:28 <dajobe> A:See [http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/2004-03-02#T08-04-28|discussion]

08:04:29 <dc_rdfig> Added comment A1.

08:07:36 <dajobe> danbri intros

08:07:54 <dajobe> Guus Schreiber

08:08:52 <dajobe> GS - co-chair webont, recently finished work

08:09:17 <dajobe> considering how to support people building semweb apps

08:09:27 <dajobe> in particular getting them on to their first steps

08:10:09 <dajobe> semantic web best practices wg (SWBP)

08:10:23 <dajobe> first idea was to merge rdf-ig and a best practices wg

08:10:50 <dajobe> but it was recognised there were two fora - a low entrance barrier chatty place, no telcons

08:11:10 <dajobe> and second a more focused groups, not producing REC-track work but pref. short notes on getting started with semweb

08:11:27 <dajobe> Considered various user types for SWBP work

08:11:31 <dajobe> language developers

08:11:36 <dajobe> those using them to make apps

08:11:39 <dajobe> and then the app users

08:11:50 <libby>http://www.w3.org/2003/12/swa/swbpd-charter.html

08:11:51 <dc_rdfig> B: http://www.w3.org/2003/12/swa/swbpd-charter.html from libby

08:12:02 <libby> B:|Semantic Web Best Practices and Deployment (SWBPD) Working Group Charter

08:12:03 <dc_rdfig> Titled item B.

08:12:17 <dajobe> also grow the rdf-ig to swig-ig growing it's scope

08:12:44 <dajobe> danbri - ontologists, rule enthusiasts have tended to live on d ifferent lists from rdf-ig

08:13:05 <RalphS>http://www.w3.org/2001/sw/BestPractices/

08:13:06 <dc_rdfig> C: http://www.w3.org/2001/sw/BestPractices/ from RalphS

08:13:07 <dajobe> now time to re-think how to structgure this

08:13:24 <RalphS> C:|Semantic Web Best Practices and Deployment Working Group home page

08:13:25 <dc_rdfig> Titled item C.

08:13:25 <dajobe> GS - IG people can work at the WG-level or taskforces (more later)

08:13:40 <dajobe> charter, as chumped B: ...

08:13:47 <dajobe> announced last week

08:14:05 <dajobe> meeting thu/fri this week here in Cannes

08:14:33 <dajobe> scope -

08:14:54 <dajobe> general aim is to support, in particular, app developers in semweb field

08:15:20 <dajobe> To do this we want to do the "low hanging fruit" - take the best practices already there.

08:15:52 <dajobe> One area is the vocabularies, ontologies, thesauari already existing, and we want to make available in public in rdf/owl

08:15:58 <dajobe> [rdf/owl = rdf or owl or both ]

08:16:22 <dajobe> people who have these thesauri have done conversions as well as writing new ones

08:16:36 <dajobe> It would be good if they were easily available to be used.

08:17:10 <dajobe> This will not be like the daml ontology library where you can put anything you want in it, but rather things tha thave acceptance in the community.

08:17:20 <dajobe> Write down what steps to take to do conversions.

08:17:37 <dajobe> Making available in public on the web.

08:18:01 <dajobe> We have a given a number of criteria for the workl.

08:18:19 <dajobe> Some WG members supporting ontology owners and helping making available.

08:18:32 <dajobe> The kinds of corpora are listed in the charter, B: above

08:21:39 <dajobe> The likely candidates for some of these things are in the public field, for govt or cultural heritage, where there is less of a business reason to do this commercially.

08:21:54 <dajobe> GS outlines more of the thesauri

08:22:28 <dajobe> Interested in a small use of vocabs that are broadly used, than a large set.

08:22:49 <dajobe> 4 criteria for an ontology to be in scope (from charter)

08:23:14 <dajobe> [[

08:23:15 <dajobe> The ontology is the result of the collaborative work in some community and represents some form of consensus.

08:23:15 <dajobe> The ontology has a user community.

08:23:16 <dajobe> The ontology owner is maintaining the ontology.

08:23:20 <dajobe> The ontology owner is willing to make the ontology publicly available on a Royalty Free basis

08:23:22 <dajobe> ]]

08:23:48 <dajobe> any other ontology is out of scope for our WG.

08:24:01 <dajobe> danbri: Do you expect to change namespaces?

08:24:16 <dajobe> ... one job I'd like is guidance for namespace owners to do the right thing

08:24:42 <dajobe> ... If I die, what guide would you give to keep this running (such as foaf) longer term?

08:24:50 <dajobe> *laughter*

08:25:08 <dajobe> GS - Hosting on W3C site. And gving advice on particular namespaces

08:25:20 <dajobe> This kind of thing is within our scope.

08:25:36 <dajobe> RalphS has spent a lot of time narrowing down the scope.

08:25:50 <dajobe> So it doesn't become something like a "mini-TAG".

08:25:54 <RalphS> [I would say 'clarifying']

08:25:56 <dajobe> Jeremy Carroll -

08:26:26 <dajobe> JC asks about endorsing (something W3C doesn't do), quality of work is near that.

08:27:16 <dajobe> GS - Quality from the charter points, for people who have used things that gives some indication the work is 'good'.

08:27:41 <dajobe> some people have condemned the wordnet in rdf approachs as bad, yet they enable apps that are entirely impossible without it.

08:28:00 <dajobe> The only stick we have is to show quality to see if it creates a user community.

08:28:09 <dajobe> s/to see/by seeing/

08:28:27 <dajobe> Second area - FAQs (1.2.2 charter)

08:28:54 <dajobe> and taskforces.

08:29:08 <dajobe> One of them is the rdf-in-html group, that will be folded into the SWBP.

08:29:54 * DanC waves from another meetign

08:29:57 <dajobe> May also include deployment issues such as XSD.

08:30:22 * JosD waves to DanC from SWIG

08:30:36 * kendallclark waves back from the same other meeting

08:30:38 <dajobe> Also may work on ontological modelling things such as doing "defaults"

08:31:06 <dajobe> Design patterns for constructing ontologies

08:31:22 <dajobe> leading to working group notes for these issues, providing guidance to the common modelling problems for app developers.

08:31:33 * DanC thinks maybe a topic change is in order

08:31:56 <dajobe> dajobe has changed the topic to: Semantic Web Interest Group meeting, Cannes, France - http://rdfig.xmlhack.com/

08:32:22 <dajobe> dajobe has changed the topic to: Semantic Web Interest Group meeting, Cannes, France http://www.w3.org/2001/sw/interest/meetings/tp2004.html - weblog http://rdfig.xmlhack.com/

08:32:47 <dajobe> GS describes more possible areas from the charter

08:32:58 <dajobe> GS -

08:33:11 <dajobe> Intend to not duplicate work (of course) and link to existing work.

08:33:29 <dajobe> Related work in US, DAML and in Europe, (?) knowlege web

08:33:47 <dajobe> which will be useful as input to the SWBP work.

08:34:53 <dajobe> (discussion of where's the money from?)

08:35:03 <JosD> Kwowledge Web is about integrating research in the Semantic Web (if I understood it well)

08:35:54 <dajobe> GS - welcome W3C fellow contributions (from members)

08:36:09 <dajobe> Third area (1.2.3 in charter): Repository of tools and demo apps

08:36:22 <dajobe> Don't think there's a big need here, except for gaps that don't already exist.

08:36:36 <dajobe> Would be v happy if this already exists.

08:36:41 <dajobe> Tools, example apps.

08:37:03 * danbri needs a scribe offer for post-coffee session, 10.30-11.00

08:37:04 <dajobe> 1.2.4 Focuse: links to related techniques

08:38:14 <JosD> Re: KnowledgeWeb some background at http://www.iti.gr/db.php/en/projects/KnowledgeWeb.html

08:38:45 <dajobe> giving advice on connecting to other fields

08:38:46 * danbri pondering FP6-phase4 project possibilities...

08:39:07 <dajobe> communtiies in software engineering

08:39:26 <dajobe> GS - we have ongoing work with OMG for work on ontologies, had 4 submissions

08:39:30 <dajobe> UML profile

08:39:47 <dajobe> using the UML symbols to represent a different language

08:40:05 <dajobe> The advantage is that people can use the same metaphors/diagrams, tools but in the background generate owl/rdf

08:40:18 <dajobe> - a number of technical problems, something should be produced.

08:40:21 <dajobe> Topic Maps

08:40:31 <dajobe> Looking for Bernard V(?)

08:40:38 <dajobe> who will be present Thu or Fri

08:40:52 <dajobe> would be good to relate things, record in a note.

08:41:12 <dajobe> danbri - SWAD europe workshop at XML europe on this topic

08:41:28 <dajobe> how can research projects feed into the standards track

08:41:34 <dajobe> GS - last is semantic web and database

08:41:49 <dajobe> (still on 1.2.4 bullets)

08:42:14 <dajobe> They are already in a relational database format, or have some XML format.

08:42:20 <dajobe> The relational tables one is actually easier

08:42:39 <dajobe> but having the XML serialization means you have to rip the semantic out again from the XML cruft

08:42:45 <dajobe> [/me paraphrases]

08:42:59 <dajobe> 1.3 Out Of Scope

08:43:09 <dajobe> GS - compromises, heh

08:43:23 <dajobe> 2 Deliverables

08:43:28 <dajobe> Typically we make WG notes.

08:43:40 <dajobe> Possibly REC track, but not expecting it.

08:44:01 <dajobe> The current schedule is through 31 Jan 2006

08:44:14 <dajobe> hope to produce 3 WG notes each 6 months plus published vocabularies.

08:44:23 <dajobe> More would be nice!

08:44:53 <dajobe> Expect a WG of 15-20 or so but those wouldn't be the only participants

08:45:12 <dajobe> the IG will be interetesed in other parts therefore we have this Task Force idea

08:45:40 <dajobe> A short ask force with some WG participants, some outside - short-term, small groups

08:46:09 <dajobe> Invited experts

08:46:58 <dajobe> danbri - experts as editors, ...

08:47:15 <dajobe> lots of work in the IG without going through an IPR hoops, just done the work

08:47:22 <dajobe> so lots of informal docs on websites, wiki

08:47:28 <dajobe> but nothing on w3c tech reports page

08:47:34 <dajobe> so want to figure out some of the problems

08:47:52 <dajobe> want to get notes published from the IG people, but it may just be invited experts

08:48:09 <dajobe> GS - If they are not WG members in the task force, they would typically come from the sw-ig

08:48:36 <dajobe> Using normal WG channels, some telcons, reach consensus to complete

08:49:02 <dajobe> The SWBP now has a home page

08:49:25 <dajobe> dc_rdfig:view 1

08:49:25 <dc_rdfig> C: Semantic Web Best Practices and Deployment Working Group home page (http://www.w3.org/2001/sw/BestPractices/)

08:49:27 <danbri> see http://www.w3.org/2001/sw/BestPractices/

08:49:33 <dajobe> telcons every 2 weeks

08:49:44 <dajobe> contribute ~20% of their time

08:50:00 <dajobe> with expectation of non-REC track work

08:50:17 <dajobe> GS - for the rest, see charter, agenda for kickoff meeting Thu/Fri

08:50:25 <dajobe> danbri - observers?

08:50:47 <dajobe> GS - everyone was granted status, but if they all show up we have a space problem :)

08:51:28 <dajobe> GS - call to participate went out friday, would actually be hard to be a member today\

08:52:06 <dajobe> discussion of timing issues re meetings, agenda

08:52:17 <dajobe> Jeremy Carroll - relationship between sw-ig and swbp re documents

08:52:29 <dajobe> the relationship is often via documents, reviewing

08:52:42 <dajobe> for swbp a task force creates a doc that the swbp reviews

08:52:49 <dajobe> and no sw-ig involvement?

08:52:52 <dajobe> danbri -

08:53:03 <dajobe> Aspects of this we are making up as we go along.

08:53:18 <dajobe> we have 627 on rdf-interest, 346 on logic, 145 on calendaring

08:53:25 <dajobe> so lots of people moderating or deleteing messages

08:53:50 <libby> gosh that's a lot of calendarers

08:54:08 <dajobe> The calendaring task force/list has existing longer than twe thought, and people drop in as they have calendaring issues, so it'll probably continue

08:54:16 <dajobe> GS

08:54:31 <dajobe> It's a kind of new idea, we will have to see how it works out.

08:54:37 <dajobe> A public WG

08:54:47 <dajobe> see no reason why the reviews cannot be open.

08:54:59 <dirkx> \who patri

08:55:18 <dajobe> JC - document comment before publication reviewed by sw-ig

08:55:47 <dajobe> GS - barried for 2nd edition of notes is lower, so easier to update

08:55:55 <dajobe> RalphS confirms, yes expect it

08:56:20 <dajobe> danbri - "Best" in the WG title is just a best-right-now, may be updated

08:56:26 <dajobe> GS

08:56:36 <dajobe> A stepwise process for conveting a thes into owl/rdf

08:56:51 <dajobe> could ge tthe 3-4 people working on it, could make somethng now

08:57:04 <dajobe> Expect in a couple of years, omething better could be done.

08:57:18 <dajobe> bijan - going to address architectural issues or avoid them?

08:57:24 <dajobe> semantic web arch issues

08:58:00 <dajobe> Have lots of diagrams of semweb techs.

08:58:14 <dajobe> "Confusing even if they make sense." - bijan parsia

08:58:27 <dajobe> for example, what is a semantic layer?

08:58:37 <dajobe> SWBP says take it to tag, tag will likely say out of scope.

08:58:42 <dajobe> GS

08:59:15 <dajobe> Indicating by linking to useful educational text, that's possible.

08:59:22 <dajobe> Not a kind of mini-semweb TAG.

08:59:29 <dajobe> danbri:

08:59:47 <dajobe> considered various group organisation/charters

09:00:00 <dajobe> GS - the sw-ig is the place for discussing architecture issues

09:00:12 <dajobe> danbri:

09:00:30 <dajobe> In a much better place for questions to go from sw-ig - to swbp, sw coordination group.

09:00:38 <dajobe> brian mcbride (bwm):

09:01:14 <dajobe> the term practice in "Best Practice" is about practice, not about updating an RDF core spec

09:01:19 <dajobe> you update it as time goes by

09:01:32 <dajobe> so it's a different thing (penny drops)

09:01:38 <dajobe> best not get into arch issues

09:01:53 <dajobe> I would like the web to have 1 architecture, have one place to do it.

09:02:21 <dajobe> bijan - for WS too?

09:02:41 <dajobe> Discussion of web/WS arch, where it's being discussed.

09:02:50 <dajobe> Charles McCathie Neville (chaals)

09:03:21 <dajobe> If an architectural discussion happens in the sw-ig, what happens to it?

09:03:46 <dajobe> danbri: you are asking really, "who" does something with synthesising info, practice, discssion summaries

09:04:01 <dajobe> we have a rough shape of a pipeline - sw-ig for rough ideas

09:04:11 <dajobe> discussions will get noticed

09:04:26 <dajobe> (some swbp people might be watching the list deliberately)

09:04:35 <dajobe> then looking for new best practtice work needs

09:04:42 <dajobe> Jeremy Carroll:

09:05:10 <dajobe> This is a theoretical problem with where to put it, might make a member submission, might do a WG note.

09:05:40 <dajobe> There is not a lot of work in the activity for making lots of new arch decisions now.

09:05:49 <dajobe> dawg for query/retrieval

09:06:01 <dajobe> we've made some progress, lets make money now not more big arch specs

09:06:20 <dajobe> chaals: if you have a great idea, no expectation for new WGs to pop up

09:06:25 <dajobe> GS: take it to sw-cg

09:06:25 <dc_rdfig> Label GS not found.

09:07:02 <dajobe> danbri: hopefully we will have made a forum where frustrated developers can get help with their semweb apps

09:07:37 <dajobe> GS - rdf/html task force

09:07:53 <dajobe> now has found it's place in a WG

09:08:01 <dajobe> Social meaning mini=task-force

09:08:12 <dajobe> fostered by sw-cg, for inter wg-ig things

09:08:36 <dajobe> danbri - agenda review

09:09:06 <dajobe> danbri summons bijan for WS/semweb lieason

09:09:22 <dajobe> Bijan Parsia - University of Maryland

09:11:04 <dajobe> work with jim hendler

09:11:13 <dajobe> on pretty much all WS / semweb groups

09:11:55 <dajobe> Semantic Web Services

09:12:01 <dajobe> intro to WS tasks

09:12:27 <dajobe>http://www.mindswap.org/~bparsia/talks/2004/march-w3c-swig.html

09:12:27 <dc_rdfig> D: http://www.mindswap.org/~bparsia/talks/2004/march-w3c-swig.html from dajobe

09:12:43 <dajobe> D:|Semantic Web Services, Bijan Parsia

09:12:43 <dc_rdfig> Titled item D.

09:12:53 <dajobe> D:written at 2am, so he apologies for that :)

09:12:54 <dc_rdfig> Added comment D1.

09:13:33 <dajobe> bijan has been running between semweb and WS wgs so far

09:13:43 <dajobe> big issues are with WSDL

09:13:53 <dajobe> but likely won't have time for them in depth

09:13:57 <danbri> Is there anyone here f2f in Cannes with IRC access who could scribe from coffee until lunch?

09:13:57 <dajobe> quite critical, need help

09:14:11 <dajobe> Slide: WSDL

09:14:28 <dajobe> has a requirement to provide a mapping to semweb/rdf

09:14:39 <dajobe> BP - joined the group this summer to shepherd that

09:15:22 <dajobe> the semweb has somewhat of a credibility gap with the WS people

09:15:34 <dajobe> need to come up with the goods

09:15:56 <dajobe> once a REC appears, we need to consume, produce these

09:16:06 <dajobe> and show the WS community the pay off

09:16:21 <dajobe> starting WD production work after this meeting

09:16:41 <dajobe> Two big issues - using rdfs/owl types for describing message content

09:16:48 <dajobe> more difficult since they removed message parts

09:16:54 <dajobe> becoming very xml schema centric

09:17:11 <dajobe> becoming owl/rdf type classes to express things about service contracts

09:17:20 <dajobe> some issues about open world and incomplete information

09:17:32 <dajobe> item 2 - has an abstract syntax

09:17:43 <dajobe> ... (too fast ) ....

09:17:59 <dajobe> might have to model xml schema compent model

09:18:15 <dajobe> RalphS: there might be wider interest?

09:18:22 <dajobe> BP - sounds like a job for best practice

09:18:25 <dajobe> GS - out of scope

09:18:34 <dajobe> BP - WS-Arch just died

09:19:14 <dajobe> arch group used owl to help their understanding

09:19:32 <dajobe> part of the published doc - first non-semweb use of an owl ontology, in WS

09:19:44 <dajobe> hope to be able to use it for mapping to the arch ontologies

09:19:56 <dajobe> that means the WS-Arch docs need to evolve, not be orphaned

09:20:03 <dajobe> hope there is a successor WS-Arch group

09:20:40 <dajobe> WS-Arch were quite enthusiastic with owl for this, took months to consider and reject UML for this

09:20:45 <dajobe> Slide WS-Choreograph

09:21:24 <dajobe> describing the how it works, not why the chor was done

09:21:56 <dajobe> Slide Semantic Web Servics IG

09:22:10 <dajobe> (BPEL - a scripting language for web services)

09:23:00 <dajobe> OWL-S and SWSI semweb services groups

09:23:08 <dajobe> they all suck at working in public

09:23:20 <dajobe> so now agreed to use the SWS IG mailing list

09:23:21 <dajobe> s

09:23:52 <RRS>http://lists.w3.org/Archives/Public/public-sws-ig/

09:23:52 <dc_rdfig> E: http://lists.w3.org/Archives/Public/public-sws-ig/ from RRS

09:23:58 <dajobe> thanks

09:24:05 <RRS> E:| Semantic Web Services Interest Group public list

09:24:06 <dc_rdfig> Titled item E.

09:24:07 <dajobe> BP - intend to take my mapping work to the sws-ig

09:24:12 <RRS> E:| Semantic Web Services Interest Group public list archive

09:24:13 <dc_rdfig> Titled item E.

09:24:24 <dajobe> Slide UDDI

09:24:50 <RRS> RRS is now known as RalphS

09:24:57 <dajobe> UDDI people come to the sws-ig to get some help

09:25:22 <dajobe> Slide OWL-S (was DAML-S)

09:25:39 <dajobe> expect to wrap up May 2004++

09:26:14 <dajobe> Maryland been working with task computing with Fujtisu of america, commercialising

09:26:32 <dajobe> broad scope

09:26:43 <dajobe> does all WS tasks from discovery ...

09:27:21 <danbri> E:Eg. [http://lists.w3.org/Archives/Public/public-sws-ig/2004Feb/0030.html|UDDI and Semantics] request for review/involvement.

09:27:21 <dc_rdfig> Added comment E1.

09:27:31 <dajobe> Slide SWSL

09:27:48 <dajobe> intended to be successor to OWL-S (on US side)

09:28:05 * dajobe assumes L=Logic

09:28:19 <dajobe> ah, no

09:28:28 <dajobe> Semantic Web Services Languages [committee]

09:28:53 <dajobe> self-hosting

09:29:10 <dajobe> partially supported by various intiiatives

09:29:35 <JanneS> http://www.daml.org/services/swsl/ - doesnẗ reply to me - maybe network prob from France

09:29:36 <dc_rdfig> F: http://www.daml.org/services/swsl/ from JanneS

09:29:58 <dajobe> BP - intention is that it moved items from standards to industry

09:30:11 <dajobe> Slide: tools etc.

09:30:18 <dajobe> stuff from Maryland

09:30:24 <dajobe>http://www.mindswap.org/

09:30:24 <dc_rdfig> G: http://www.mindswap.org/ from dajobe

09:30:57 <dajobe> F:|Semantic Web Services Language (SWSL) Committee

09:30:57 <dc_rdfig> Titled item F.

09:30:58 <dajobe> Statement of Mission and Objectives

09:31:00 <dajobe> eek

09:31:07 <dajobe> [url works for me]

09:31:38 <dajobe> G:|Semantic Web Research Group, University of Maryland

09:31:39 <dc_rdfig> Titled item G.

09:32:03 <dajobe> BP outlines the great stuff being done at his org

09:32:18 <dajobe> working on photo markup, particularly sub regions

09:32:23 <dajobe> java app

09:33:10 <dajobe> Issues

09:33:16 <dajobe> They are all very hard, give me money.

09:33:36 <dajobe> Automated composition - seems to be AI planning alike

09:34:00 <dajobe> WS tends to avoid talking about sate - reasoning does

09:34:06 <dajobe> state

09:34:30 <dajobe> The Decker Problems - no time now, see sws-ig

09:34:47 <dajobe> Very hard to express lots of things about services in a restricted lang like OWL, OWL-S, ...

09:35:33 <dajobe> Need at least rules, probably more

09:35:42 <dajobe> WS could be an attractor for rules

09:35:53 <dajobe> particularly with talking about polices expressed as rules

09:36:03 <dajobe> *applause*

09:36:08 <dajobe> bijan ends

09:36:18 <dajobe> coffe break till 11:06 GMT+1

09:36:24 <Remosi> Mmm Coffee.

09:46:31 <arnarl> hi

09:47:24 <Remosi> hi

10:15:14 * bwm wonders if there is a scribe

10:15:33 <bwm> conceptual graph based search

10:15:42 <danbri> corby - CGs are a kind of semantic net

10:15:44 <bwm> contecptual graph from Sowa in '80s

10:15:53 * danbri can scribble for a bit

10:15:58 * bwm fine

10:16:01 <jose-tp> (Talk by Olivier Cohu, INRIA Sophia Acacia team)

10:16:03 <danbri> (we can get slides, no need reproduce them)

10:16:28 <danbri> (shows RDF intro slide)

10:16:42 <RalphS> s/Cohu/Corby/

10:17:37 <RalphS> Olivier's project is Corese

10:17:42 <danbri> hmm irc problems

10:17:55 <danbri> see http://www-sop.inria.fr/acacia/personnel/corby/top.html re olivier

10:18:17 <danbri> ...represents CG engine as an RDF system to outside code

10:18:18 <RalphS>http://www-sop.inria.fr/acacia/soft/corese/

10:18:19 <dc_rdfig> H: http://www-sop.inria.fr/acacia/soft/corese/ from RalphS

10:18:22 <RalphS> H:| Corese : an RDF engine based on Conceptual Graphs

10:18:23 <dc_rdfig> Titled item H.

10:18:31 <danbri> ...using query patterns inside RDF graphs

10:18:46 <danbri> (appear to be strings, eg ~, ? etc whre URIs expected?)

10:19:12 <danbri> [JosD, are you happy doing your Euler talk after the Contexts discussion?]

10:19:19 * DanC looks around for timbl...

10:19:26 <danbri> ...implements Approximate Projection

10:19:35 <JosD> no problem DanBri :)

10:19:42 <danbri> bon :)

10:19:54 <danbri> eg. [TechReport] (author) [Person]

10:20:00 <danbri> vs [ResearchReport] (author) [Person]

10:20:10 <danbri> ...notion of proximity in contex of the rdfs

10:20:20 <danbri> ...compute ontological distance

10:21:24 <danbri> ...query lang operators with variables

10:21:32 <danbri> ...small constrain solver within graph projection

10:21:36 <danbri> ...type operators

10:21:48 <danbri> ...can retrieve docs, type of which strictly equal to TechReport

10:22:00 <danbri> ...or inferred

10:22:12 <danbri> ....can retrieve graph pattern with variable length paths

10:22:22 <danbri> eg. *3

10:22:38 <danbri> ...when you don't know / master completely the rdf schema...

10:22:48 <danbri> ...you can specify that you want the answer for all the paths

10:23:05 <danbri> bwm: path length, does it have to be a fixed constant

10:23:16 <danbri> olivier: ...yes, needs to be a fixed constant

10:23:23 <danbri> ...have a generic property

10:23:34 <danbri> (a top of rdfs:subPropertyOf hierarchy?)

10:23:41 <danbri> ...distinct, update etc too

10:23:49 <danbri> ...starting to implement bits of OWL

10:23:55 <danbri> ...transitive, symmetric, inverseOf, ...

10:24:01 <danbri> ...inference rule engine

10:24:20 <danbri> ...if/then stuff

10:24:26 <danbri> ...underlying model is a graph model

10:24:49 <danbri> ...possible inferences are refining concept type, eg. a person is declared to be a participant

10:24:53 * bijan wonders what's going on

10:25:15 <danbri> (lightning talk from Olivier Corby)

10:25:16 <bijan> fixed constant length path? Of what?

10:25:21 <danbri> ...also have a SW Server

10:25:32 <danbri> ...corese engine integrated via tomcat

10:25:38 <danbri> ...query via html forms/jsp etc

10:26:00 <danbri> ...applications: about 8 apps within EU projects for corporate memory, KM platform, etc.

10:26:13 * danbri wonders what licensing terms are for Corese

10:26:31 <danbri> ...quick demo

10:26:35 <danbri> ...SW server

10:26:59 <danbri> ...[shows rdfs being navigated (via xslt styling), visiting instances ...showing all properties of a doc]

10:27:23 <danbri> ..[table of terms, definitions, ...]

10:27:47 <danbri> ...can ask for docs known within system, or in terms of subclassing

10:27:49 <danbri> ...etc

10:28:42 <danbri> (anyone capture that Q?)

10:29:06 <JosD> why don't you use XQuery?

10:29:11 <danbri> jjc: you said it was based on CGs...

10:29:25 <danbri> ...is this the older INRIA engine... same issues/features? (redundancy...)

10:29:43 <danbri> o: we used some canadian thing

10:30:16 <danbri> jjc: INRIA has been working on CGs for a decade or more, r u building on that?

10:30:23 <danbri> o: yes (I think)

10:30:29 <danbri> danbri: what are licensing terms?

10:30:38 <danbri> o: free for academic/nonprofit, but not opensource

10:30:48 <danbri> ?: interesting to see how close CGs, RDF are

10:31:02 <danbri> ...sicne CGs have been extended in many ways, eg. nested graphs for contexts

10:31:10 <danbri> ...rules, algorithmic optimisations...

10:31:18 <danbri> ...lots of tools that have been developed

10:31:37 <danbri> jjc: a while ago i did some work on graph isomorphism

10:31:44 <danbri> ...jerome pointed out CGs

10:32:05 <danbri> ?(speaker?): I should post a small bibliography to the RDF list

10:32:12 <danbri> ...ICCS conf on conceptual structures

10:32:43 <jose-tp> ? == Jean-Francois Baget, INRIA Rhone-Alpes

10:33:43 * JosD semWaves to TimBL

10:34:44 * timbl waves back sleepily

10:35:36 <danbri> jjc: presenting Modelling Context using Named Graphs

10:35:46 <danbri> ...work with Chris Bizer, related to trix work w/ PatrickS too

10:35:57 <danbri> [slides will have details] @@url?

10:36:11 <danbri> ..."the context of a statement is its surrounding graph"

10:36:26 <danbri> [slides look rich, am not going to do detail]

10:36:36 * nmg wonders who's presenting at the moment

10:37:08 <danbri> jeremy carroll:(jjc) which of these graphs are asserted?

10:37:20 <danbri> ...also de re / dicto... isssues w/ rdf reification

10:37:39 <danbri> jjc: chris claimed all uris must have a single interpretation

10:37:47 <danbri> [i don't think tag or ietf agree on that --danbri aside]

10:38:01 <danbri> jjc: [slide 4 shows a named graph in sysntax]

10:38:04 * danbri wonders if ericp here

10:38:09 * bijan notes that that's nonesense (about the uris)

10:38:21 <danbri> jjc: TrQL (extension to RDQL for named graphs)

10:38:33 <danbri> example: [handy; missed]

10:38:42 <danbri> ...data syndication: provenacne tracking

10:38:55 <danbri> ...eg named graphs w/ prov info

10:39:00 <danbri> ...g1 talks about g2 etc

10:39:07 * timbl can't follow from the irc though it sounds fascinating

10:39:26 <Remosi> timbl: it's easier when there are slides to read along with

10:39:31 * nmg agrees

10:39:33 <danbri> ...signing rdf graphs

10:39:34 <Remosi> however, I don't think we found slides for this talk

10:39:45 <dajobe> things like G1 (triple1, triple triple) G2 (G1 thing thing)

10:39:51 <danbri> [can someone get the slide url from www-archive ?]

10:39:52 <dajobe> where G1, G2 are graph names

10:39:58 <dajobe> it's not there in www-archive

10:40:08 <danbri> [ok we get later]

10:40:25 * timbl What slides not on the web - and this is a w3c ig? 0.5 ;-)

10:40:26 <danbri> ...signing need a crpyto module [vocab?]

10:40:32 <bijan> Do I need to paste in my slides uri?

10:40:41 <dajobe> bijan: no, I found them

10:40:41 * danbri notes ppl are writing slides shortly b4 talks

10:40:46 <danbri> bijan, yes please!

10:40:51 <bijan> Thanks dajobe

10:40:56 <dajobe> see chump, url there already

10:40:56 <bijan> Er..

10:41:27 <danbri> jjc: scoping assertions: eg. of different notions

10:41:30 <nmg> is there a ref for the "TrQL" that jeremy mentioned?

10:41:37 <JosD> slides are at http://lists.w3.org/Archives/Public/www-archive/2004Feb/att-0079/Overview.html

10:41:42 <danbri> thx!

10:41:55 <JosD> now at slide 9

10:42:44 <JosD> now at slide 10

10:42:50 <libby>http://lists.w3.org/Archives/Public/www-archive/2004Feb/att-0079/Overview.html

10:42:50 <dc_rdfig> I: http://lists.w3.org/Archives/Public/www-archive/2004Feb/att-0079/Overview.html from libby

10:42:52 <danbri> ...liar paradox: 'this sentence is false' using log:implies

10:43:15 <danbri> ...maybe log:implies is wrong/problematic

10:43:16 <libby> I:|Modelling Context using Named Graphs by Chris Bizer & Jeremy J. Carroll

10:43:17 <dc_rdfig> Titled item I.

10:43:19 <danbri> JosD: pls elaborate!

10:43:33 <danbri> jjc: [oh dear danbri not capturing this properly]

10:43:55 <danbri> jjc: ... including the metatheory in with the theory, is a known dubious practice for mathmatics

10:44:06 <dajobe> jjc beats up log:implies

10:44:15 <danbri> ...a diff way of thinking about log implies... rather than a semantic consequence, think of it as a theory

10:44:22 <danbri> ...lots of math foundation work on this distinction

10:44:24 * timbl sighs ... wishes he were there

10:44:31 <danbri> ...log:implies glosses over that known hard problem

10:44:44 <danbri> can someone take over scribing please? bwm?

10:44:51 * danbri wants to try ichat/videeo for timbl

10:44:54 <danbri> timbl, got yr mac?

10:44:59 <timbl> Yup

10:45:00 <danbri> jjc: de re / dicto

10:45:48 <danbri> (gives example)

10:46:17 <dajobe> Slide: De Re or De Dicto

10:46:26 <RalphS> jjc: De Dicto is a direct quote

10:46:27 <bijan> OOOH

10:46:31 * timbl thinks lide 11 seems to be in favor of de dicto

10:46:38 <bijan> I wish I were over there

10:46:40 <danbri> timbl, aim/ichat danbri2002

10:46:43 * bijan loves de re/de dicto

10:46:57 * JosD believes however that log:implies is quite OK, actually almost the only thing we need :)

10:47:01 * danbri would like chat about making a propositional attitudes namespaces, different designs for that

10:47:35 <dajobe> jjc: if we assume de dicto, and want to do a merge, we have probs

10:47:57 <dajobe> if they use the same uris, we need to judge who made uris correctly

10:48:06 <dajobe> who do we trust to even use the same language as u

10:48:07 <dajobe> s

10:48:31 <dajobe> jjc: if we really want de dicto, can use the xsd URI datatype

10:48:46 <dajobe> danny j weitzner;pushs a question

10:48:58 <dajobe> jjc: "log:implies is broken, timbl"

10:49:13 <dajobe> slide Some other use cases

10:49:20 <nmg> is jjc talking about using the same URIs to mean different things, or different URIs to mean the same thing?

10:49:39 * dajobe dunno, this talk is 10,000m above my head

10:49:56 * JosD thinks that the misunderstanding is maybe that the premise subject of log:implies is not asserted...

10:49:58 <dajobe> slide Beliefs

10:49:59 * nmg regrets not being in Cannes

10:50:05 <dajobe> not made muchb progerss

10:50:10 <dajobe> slide Trust Evaluations

10:50:20 <dajobe> over to Chris Bizer

10:50:33 <dajobe> bwm asks - belief things brings the de re/de dicto in tofocu

10:50:34 <dajobe> s

10:50:37 <dajobe> ... superman example

10:50:39 * dajobe sighs

10:50:50 <dajobe> jjc reminds ppl

10:51:13 * timbl has been though all thos but can't pick up what jjc waas just arguring

10:51:30 <dajobe> jjc: beliefs we thing of as quoted

10:51:33 <dajobe> think of

10:51:45 <dajobe> maybe the shared conceptualization breaks down there

10:51:50 <dajobe> chis bizer

10:51:56 <timbl> In { log:implies {} tyhey are quoted.

10:52:04 <gk> bwm: lois believes that superman can fly; we know superman==clark kebnt; therefore, does Lois believe clark kent can fly?

10:52:04 <danbri> tim, you hear ok?

10:52:05 <timbl> This is good.

10:52:09 <dajobe> maybe we should project the irc

10:52:42 * danbri using isite's mic

10:52:50 <dajobe> the projected slides are different here from what's in ww-archive

10:52:56 * bijan notes that the BOF table sign up is up

10:53:05 * bijan would be interested in a couple of BOFs

10:53:08 <dajobe> slide Example Trust Policies

10:53:17 * dajobe proposes the Bijan BOF

10:53:19 <bijan> WSDL RDF mapping bof would be fine

10:53:32 <bijan> And useful, perhaps

10:53:39 * bijan proposes the dajobe bash

10:54:17 * bijan wishes he was in cannes too

10:54:30 <Remosi> is anyone still scribing?

10:54:34 <dajobe> cb - advertises workshop on trust at (?) conf by Jennifer Golbeck, Maryland

10:54:39 <dajobe> jjc

10:54:50 * bijan meant to put that in his slides; forgot

10:55:18 <gk> Bijan: I wanted to ask you about your request for tools work re. WSDL/RDF interworking ... this might be interesting for me, but I'm not sure where to pick up the threads.

10:55:40 <dajobe> jjc: the assertivness layer of the graph is more moving into the higher layers rather than syntactic

10:55:44 <dajobe> Issue List

10:55:52 <timbl> There always has to be a chain of things leading one to belive a graph.

10:55:53 <dajobe> stops

10:55:55 <yonderboy> .google hecklebot

10:55:57 <datum> hecklebot: http://joi.ito.com/joiwiki/HeckleBot

10:56:06 <bijan> gk: good, we should chat

10:56:30 <dajobe> brian mcbride

10:56:47 <dajobe> there's a relationship between a triple & graph it is in/? can model in rdf?

10:56:49 <dajobe> jjc: yes

10:57:01 <dajobe> chris did quintuples - statment & graph id, dropped that

10:57:03 <gk> Bijan, OK.

10:57:34 <dajobe> bwm: an identifier for a graph with a single triple, is not a ... lost it

10:57:39 <timbl> Nice room.

10:57:42 <dajobe> listen to the mp3s ;)

10:57:49 <danbri> huge room

10:57:55 <RalphS> bwm: .. is not an identifier for the triple

10:58:12 <dajobe> reto

10:58:16 * timbl waves back to dajobe

10:58:22 <RalphS> bwm: there are statements that will be true of one that are not true of the other

10:58:24 <dajobe> reificiatnion is not swatchable for quoting

10:58:34 <dajobe> jjc: no, I cant explain this, looks for pat hayes

10:58:53 * bijan sighs

10:59:00 <bijan> I'm not pat, but I could play him on irc ;)

11:00:42 <dajobe> jjc & bwm try to explain to reto

11:01:35 * PStickler wonders if the superman problem relates to the fact that RDF has denotational rather than intensional interpretation of URIs... ???

11:01:53 <bijan> Er...you mean extensional?

11:02:18 <PStickler> Dunno... Maybe... ;-)

11:02:35 <bijan> RDF doesn't have any intensional contexts. That is, everywhere in an RDF, you can substitued equals for equals and presever truth value

11:02:37 <dajobe> danbri mentions philosophical attitudes

11:02:50 <bijan> In other words, if you have a triple s p o.

11:02:50 <PStickler> Er, actually, no. I meant intensional.

11:02:54 <dajobe> .. from the foaf experiment

11:03:24 <dajobe> danbri - using foaf, tracking provenance, using sigs, with pgp

11:03:28 <bijan> You can replace s, p, or o with a same denoting alternative expression and you preserve the truth value of the triple

11:03:34 <bijan> Er..

11:03:35 <areggiori>http://lists.w3.org/Archives/Public/www-rdf-interest/2004Feb/0209.html

11:03:35 <dc_rdfig> J: http://lists.w3.org/Archives/Public/www-rdf-interest/2004Feb/0209.html from areggiori

11:03:36 <dajobe> ... having done that, should re-expose out - hard

11:03:42 <bijan> PStickler: huh?

11:03:56 <PStickler> That was my point. That RDF doesn't have intensional contexts, so the URI Lois uses to denote CKent/Superman doesn't reflect her perception/beliefs/knowledge about CKent/Superman in terms of one role vs. the other.

11:04:02 <areggiori> J:anohter proposal/solution to modeling contexts/scopes in RDF/XML syntax

11:04:02 <dc_rdfig> Added comment J1.

11:04:03 <dajobe> ... problem with rdf's reification vocab, can't re-syndicate it

11:04:31 <Yoshio> q?

11:04:46 <dajobe> danbri ... in XFN have uncle, which is not cross-culturally definable

11:04:53 <bijan> Right. So the contrast class isn't "denotation" but "extensional" :)

11:04:55 <areggiori> J:|contexts using RDF/XML

11:04:55 <dc_rdfig> Titled item J.

11:05:14 <bijan> You can give intensional contexts a model theory (typically using possible world semantics)

11:05:51 <bijan> It does seem that URIs could be rigid designators, which might add extra stickiness

11:05:57 <areggiori> j:where context = graph label = RDF resource i.e. simple a name (resource) representing the graph you are talking about

11:05:59 <PStickler> A similar problem is the Venus vs. Evening Star vs. Morning Star; the choice of name reflects a perception. But all three names can denote the same thing (the planet). Yet statements made using one name may not be considered true using the other name.

11:06:05 <areggiori> J:where context = graph label = RDF resource i.e. simple a name (resource) representing the graph you are talking about

11:06:06 <dc_rdfig> Added comment J2.

11:06:43 <bijan> s/may not be considered/may not be/

11:08:43 <areggiori> J:then you can stack on top of that your layers e.g. trust, provenance, "clever" inferencing by simply rdf:Describing them

11:08:43 <dc_rdfig> Added comment J3.

11:09:21 <PStickler> If you relate the different names (URIs) using owl:sameAs then statements could have free variation between the URIs and be true yet not reflect what people believe/percieve. E.g. { ex:Superman owl:sameAs ex:ClarkKent . ex:Superman ex:can ex:fly } log:implies { ex:ClarkKent ex:can ex:fly } yet Lois would not necessarily agree with the latter assertion (in absence of the basis for the inference).

11:09:53 <bijan> Yes

11:10:01 * bijan confused; didn't he say that earlier? :)

11:11:27 <RalphS> DanBri: de re concerns what you've said, de dicto concerns how you've said it

11:11:30 <bijan> Belief (typically) is an intensional context. So substituting extensional equivalents into such contexts is not (necessarily) truth preserving

11:11:59 <PStickler> Right.

11:12:21 <bijan> Lunch!

11:12:25 <RalphS> ChrisB: the de re/de dicto discussion has to do with whether there is really shared conceptualization

11:12:32 <PStickler> But RDF doesn't provide for intensional contexts, so .....

11:13:48 <PStickler> Actually, I take that 'Right' back. If it is true that superman can fly, and clark kent is superman, then it is true that clark kent can fly, even if Lois doesn't believe it because she doesn't equate superman with clark kent.

11:14:51 <timbl> DanBri, re "range" ... the info which tells a computer what to belive is capturs nowin command lines, and makefiles, and n3 rules using log:includes

11:15:01 <danbri> [swig lunch break in 17 mins. we have JosD on Euler befre then...]

11:15:15 <danbri> er [15 mins]

11:15:49 <danbri> danny: de re with context is de dicto?

11:15:56 <timbl> You don't have to make predictaes special, JJC - you do need a social convention that they have a common meaning. Adn you tell your agent to belive pole who have agreed to use the same meaning for properties.

11:16:11 <timbl> timbl is now known as timbl-deDicto

11:16:55 <JosD>http://www.agfa.com/w3c/Talks/2004/03swig/Overview.html

11:16:55 <dc_rdfig> K: http://www.agfa.com/w3c/Talks/2004/03swig/Overview.html from JosD

11:17:09 <JosD> K:|SWIG implementation experience in Euler

11:17:10 <dc_rdfig> Titled item K.

11:17:57 * JosD remembers that dere with context = de dicto

11:18:02 <gk> Thinks... (re. Lois & Superman) ... another property "isAlsoKnownAs" ? (I'm wary of using owl:sameAs for this)

11:19:08 <dajobe> JosD

11:19:23 <dajobe> Slide Running Code

11:19:25 <dajobe> Euler

11:19:28 <gk> (Then the problem can me modelled semantically by differing property extensions, while preserving URI denotations across graph interpretations)

11:19:46 <dajobe> ... we believe test casea re very important

11:20:00 <dajobe> .. explicit assumptions - explicit v imporatnt

11:20:16 <dajobe> (box at bottom of slide 2)

11:20:18 <PStickler> URIs denote resources, irrespective of perception, belief, etc. So if both ex:ClarkKent and ex:Superman both denote the same person (resource) then owl:sameAs should be correct.

11:20:52 <dajobe> EulerSharp - backwards chaning reasoner enhanced with Euler path detection

11:21:01 <dajobe> (found by Euler the person)

11:21:18 <dajobe> have provved it is complete and sound

11:21:44 <dajobe> will tell you if a given set of facts+rules supports a conclusion, i.e. very appropirate ofr test cases

11:21:57 <dajobe> use as much as possible of SWAP stuff from timbl, danc, et al

11:22:16 <dajobe> expect another release soon

11:22:40 <dajobe> (refers to DIG, not sure what it is)

11:22:50 <dajobe> small and fast :)

11:22:55 <dajobe> slide Theories

11:23:04 <dajobe> links to the rules files

11:23:14 <dajobe> shows rdfs-rules.n3

11:23:22 <dajobe> facts consitent with the rdfs model theory

11:23:26 <dajobe> then ifnerence rules for rdfs according to n3

11:23:32 <dajobe> roughly stating what the mt states

11:23:44 <dajobe> basic derivations =>

11:24:14 <dajobe> only 1 inconsitency check in rdf, w.r.t. xml literals

11:24:25 <dajobe> shows owl-rules ... more elaborate

11:24:29 <dajobe> about 150 implications

11:24:37 <dajobe> shows inconsistencies

11:25:15 <dajobe> at ### inconsistency detections

11:25:27 <dajobe> reads the "owl20i2" one

11:25:35 <dajobe> nothing can belong to an empty class

11:25:48 <dajobe> example of how it's done for owl tet cases

11:25:55 <dajobe> consistency tests can't be done this way

11:26:00 <dajobe> Slide Testcases

11:26:17 <dajobe> shows etc1.n3

11:26:22 <dajobe> derived from rdfcore manifest

11:26:31 * timbl-deDicto slide 4/5

11:26:37 <dajobe> shows etc3 results - slide 4/5

11:27:09 <dajobe> shows various examples of , say calculating easter

11:27:19 <dajobe> the result output contains the proof

11:27:29 <dajobe> slide 5 conclusions

11:27:43 <dajobe> bottm up approach,l simple first

11:28:19 <dajobe> moving to an abstracted form from hardware to software (cpu->engines, ...)

11:28:50 <dajobe> better to use a generic "semantic web cpu"

11:29:02 <dajobe> and write the software program (aka rules here) for it

11:29:06 <dajobe> just an idea

11:29:16 <dajobe> extremel pormising for automatic software generation

11:29:22 <dajobe> </josD>

11:29:36 <dajobe> Graham Klyne

11:30:20 <dajobe> noticed there are bits of rdf can't be expressed in rules, assigning the same bnode to literal?

11:30:27 <dajobe> JosD: didn't find that problem

11:30:37 <dajobe> ... pleas emake a atest case? :)

11:31:01 <dajobe> danbri -

11:31:08 <dajobe> 2nding importance of test cases

11:31:37 <timbl-deDicto> Bit of RDF which can't be expressed in rules, GK? please cc me on the example.

11:31:47 <dajobe> GK/jjc offered to do a test caes

11:32:17 <dajobe> danbri - rules & query langs - comparison?

11:32:29 <dajobe> josd - will participate in DAWG, will give evidence

11:33:48 <dajobe> jjc - declarative langs replacing computers is optomistic :)

11:34:15 <dajobe> ... useful for certain tasks. Making it all declarative won't work.

11:34:26 <dajobe> JosD: sure

11:35:55 <dajobe> danbri closes session

11:36:00 <dajobe> back 13:45 GMT+1

11:41:08 <timbl-deDicto> timbl-deDicto is now known as timbl_break

12:05:50 <grault> grault is now known as earle

12:07:12 <earle> zool: if your timezone allows you to be around before 6pm UTC, give me a shout.

12:38:50 <kaickul> hi everyone

12:43:34 <k__> k__ is now known as kclark

13:02:06 <dajobe> we reconvene, but i'm not typing again

13:02:24 <Davey> Davey is now known as D[a]vey

13:03:38 <libby> ---discussion of good/bad aspects of miling lists and other tools

13:03:53 <timbl_break> timbl_break is now known as timbl-tag

13:04:13 <libby> danny ayers: rdfig mailing list goes very quickly to technical aspects. foaf lists is fresher

13:04:33 <libby> ....evangelism angle for newbies?

13:04:57 <libby> danbri: faqs - static answers - wiki (from swad-e project)

13:05:04 <libby> ...more of these - BP group?

13:06:04 <libby> chaalsNCE: rdfig mailing list - people aren't treated badly on the list - but the list can look tech and frighhtening - no culture of asking dumb qustions and people don;t necessarily understand the fieldf enough to use FAQs

13:06:35 <libby> ....another list? change in culture? place to keep on asking stupid questions

13:06:48 <libby> ...and that community answering them when they know

13:07:22 <libby> danbri: with previous list tried to pick up on good bits of xml-dev but not the 'recreational hostility'

13:07:40 <libby> ...as chair try to get people to tone down nastyness if occurs

13:08:20 <libby> chaalsNCE: not a bullying culture on the list - this has worked. consutructive and helpful disscussion but which can be very far above you as new person

13:09:18 <libby> danbri: sometimes people treat specs as sacrosanct and blame themselves when they can;t make them work together. we could be more upfront that they do not necessrily mesh perfectly

13:09:25 <gk> (I wonder if Chaals' point is a symptom of success... RDFIG has lots of detailed "up there" stuff to discuss

13:09:49 <libby> patrickS: spcific mailing list for general users and montiored by the BP group?

13:10:09 <libby> dannyAyers: 'rdf-users'?

13:10:57 <libby> danbri: tried to broaden into weblogs - like leigh dodds' xml-deviant description - summarising threads, big picture

13:11:27 <libby> ....rise of weblogging means that many people could do this - e.g. dannyAyers with the links :)

13:11:42 <libby> ...also technoology like RSS - for lists

13:12:15 <libby> 10? 12 own weblogs, a few more read them

13:12:33 <libby> (about 25-30 people in teh room)

13:13:13 <chaalsNCE> I saw virtually the same hands for bloggers and blog-readers (a couple of readers who don't have one)

13:13:20 <libby> dannyAyers: atom - aim of rss sans politics - tried to run a weblog following those develeopments - but v difficult - weblog, other weblogs., mailing lists - would take more than one person to keep track of RDF

13:13:35 <libby> danbri: not just RDf now...

13:13:59 <libby> dajobe: planet.rdfweb.com - aggregation

13:15:08 <libby>http://planet.rdfhack.com/

13:15:08 <dc_rdfig> L: http://planet.rdfhack.com/ from libby

13:15:21 <libby> L:|planet RDF

13:15:21 <dc_rdfig> Titled item L.

13:15:28 <libby> (demo of planet rdf)

13:15:53 <libby> danbri: also feedster.com (a rss search engine) - everythign written in the weblog world about these topics

13:16:15 <libby> dajobe: need someone to be paid to do rdf-deviant

13:16:53 <libby> danbri: yep, and not a lot of editorial work form me as chair

13:18:12 <libby> philtetlow: I'm working in the real world in the users sense - SW requires a certain amount of tech competence to do this stuff; in commercial world, people who make tech decisions are not technical. but they know the value of the semnatic web. grassroots uk and further - "I understand that it's useful but not how to use it"

13:18:51 <libby> ...people are watching you - non-techinical people - and changing peoples' perceptions would be a good thing to do. newbie stuff - yep

13:19:11 <libby> danbri: were the rdf primer documents useful? (yes - a few people)

13:19:34 <libby> chaalsNCE: it is a big improvment but still a very techinical document

13:19:44 <libby> ...e.g. several syntaxes is confusing

13:20:02 <libby> ...big step in the right direction

13:20:33 <JosD> there is an [RDF Primer | http://www.w3.org/TR/2004/REC-rdf-primer-20040210/]

13:20:40 <libby> jeremy: karl dubost is doing the french translation - he has editorial comments

13:21:34 <libby> danbri: irc chats - this channel - some peopel use that a lot, (and the blog) rather than the mailing lists - instant feedback

13:21:38 <JosD> ... and an [ OWL Guide | http://www.w3.org/TR/2004/REC-owl-guide-20040210/]

13:22:04 <libby> ...18 months ago started having scheduledtopicchats - calendar,. query testcases, geo, show and tells

13:22:21 <PStickler> Not everyone has the ability to use IRC due to firewall restrictions...

13:22:43 <libby> 10-12 people use irc, fewer have been to scheduled chats

13:22:55 <dirkx> Plus emails are generally a bit more context full. :-) IMHO.

13:23:31 <kao> when counting dont forget the "lurkers" like myself :-)

13:24:02 <libby> graham: irc not good for people with no broadband or different timezones

13:24:19 <libby> jeremy: send announcements and minutes to the mailing list

13:24:26 <libby> [we did this for calendaring]

13:24:44 <libby> chaalsNCE: logging and chumping is very important - otherwise no persistence/history

13:25:18 <libby> dirkx: irc is rather fragmented compared with irc - more context

13:25:57 <libby> ...irc groups can fix small bugs quicker, but email groups seem to have more of an overview - architectural questions

13:26:34 <libby> danbri: we are an interest group not a wg - we don;t have to do any work at all;) but it would be nioce to reach a common understanding on these architectural issues

13:26:46 <libby> ...charter mentions possibility of doing audio telecons

13:27:11 <libby> ...but we're very scattered - phoning USA may not be possible for all - exclusionary

13:27:11 <dirkx> note to self: obviously more of a tracking/to-do focus of the various Bot's would go a long way to solve some of this 'fragemented' and 'dropped thread' issues.

13:27:38 <libby> ...would like to type less and talk more

13:28:56 <libby> yoshiyo: from the japanese point of view, who have english as second language, audio can preclude japanese involvement; also IRC has some tendency because of the time difference - prefers mailing lists for these reasons

13:29:10 <libby> danbri - might be interested in #sw-jp

13:29:36 <libby> yoshiso - might be a barrier between the channels?

13:30:25 <libby> danbri: interesting experiment - kota translated one of these image meetings to the japanese channel

13:30:30 * PStickler wonders if combining IRC and Babbelfish like technology would allow for real-time relatively intelligible multilingual chat...

13:31:02 <libby> tomcroucher - an occasional telecon with focussed topic might work, but nothing less focussed, wasteful

13:31:19 <libby> jeremy - prepared discussion discussed briefly in WG - less structured can be a waste of time

13:31:29 <libby> that would be very cool patrick

13:32:17 <libby> brianmcbride: from the more formal meetings on irc - less floor ocntrol than on phone

13:32:32 <kaickul> is there a rdf/semantic web application that does something useful without prior knowledge of schemas/ontologies ?

13:32:54 <libby> [??scribe brainfarts and forgets name of speaker] orderingh threads in irc would be useful

13:33:03 <libby> tomc: technically possible

13:33:25 <sh1m|cannes> jose

13:33:31 <libby> chaalsNCE: talking for an hour and listening is very hard

13:33:36 <libby> thanks sh1m

13:33:40 <libby> ??==jose

13:33:58 <libby> chaalsNCE: irc+call for non-english speakers makes it easier

13:34:03 <mdupont> hi JosD

13:34:17 <PStickler> Email also helps those who type sloooowly.. ;-)

13:34:22 * JosD hi Mike !!!

13:34:28 <libby> kaikul, mdupont, we're scrivbing from a face to face meeting - sorry for the noise

13:35:13 <libby> PStickler: I like irc - but in the RDFCore WG, if you're not using IRC you may get excluided from discussion

13:35:15 <mdupont> libby, hey, i just saw from my new strawfeed to the rdfig rss feed that jos had published some more stuff on euler and swig

13:35:27 <libby> yes he did a talk here

13:35:50 <libby> PStickler: using irc for queuing for example is tricky in that case

13:36:18 <RalphS> you don't need irc to put yourself on the speaker queue for a W3C meeting

13:36:25 <kao> kaickul: well, several people are workign on fetching schemas for things not previously known

13:36:33 <libby> danbri: would like to go beyond mailing list but unsure how to use these other tools effectively

13:37:10 <libby> tomcroucher: nntp can make it easier to respond and ocontrol the volumne of information

13:37:23 <libby> danbri: we do have rss feeds for the miling lists

13:37:43 <Wack> 14:30 * f~PStickler wonders if combining IRC and Babbelfish like technology would allow for real-time relatively intelligible multilingual chat... <- my IRC bouncer (used to?) does that

13:37:46 <libby> danbri: voice over ip e.g. vonage are becoming more common

13:37:53 <Wack> but only a few languages to and fron english

13:38:02 <libby> sounds need wack

13:38:07 <PStickler> The problem with queuing is that for those who have IRC, they can quickly jump onto the queue without interrupting the speaker, but others have to wait for a covenient moment to request that they be put on the queue, thus non-IRC folks end up often not having time to present their comments...

13:38:31 <libby> danbri: also apple users can use AIM videoconferencing etc, or audio - proprietory but useful

13:38:40 <libby> ...keep an eye on these technologies

13:39:01 <libby> dannyayers: how is the wiki doing?

13:39:16 <libby> danbri: can be messy; lots of good stuff on there

13:39:35 <libby> gk: can work well on the amiling list - send them the uri of a wiki topic

13:39:50 <libby> danbri: complements the chump - can;t change the next day

13:40:16 <Yoshio> I'd like to add the time zone problem to what I've told concerning the irc and audio conference above.

13:40:20 <earle> afternoon all

13:40:23 <libby> ----lighening talks

13:40:29 <libby> hi earle

13:40:32 <kaickul> kao: got any example ?

13:40:38 <libby> danny ayers up first

13:40:40 <earle> howdy libby

13:40:55 <libby> [we are scribign http://www.w3.org/2001/sw/interest/meetings/tp2004.html]

13:41:04 <earle> ah, is the /topic being summarised live then? I'll shut up...

13:41:14 <libby> yeah sorry for the noise earle

13:41:22 <earle> no, that's fine :)

13:41:31 <earle> I have a question but it can wait

13:41:45 <libby> #foaf no good for it?

13:42:00 <earle> Hmm. Well, it's less FOAF, more just RDF.

13:42:06 <libby> DannyAyers: XOW - Xhtml outlines and the W6 vocabulary

13:42:23 <libby> W6 vocab: who what why where how

13:42:56 <libby>http://semtext.org/2004-02/slides/

13:42:56 <dc_rdfig> M: http://semtext.org/2004-02/slides/ from libby

13:43:20 <libby> ....spends a lot of time in rss etc communites where keeping it simple is important - but can become too simple

13:43:30 <libby> M:|XOW - Xhtml outlines and the W6 vocabulary

13:43:30 <dc_rdfig> Titled item M.

13:44:14 <libby> ....the users want features - moving towards production of RDF to the users - not RDF/XML

13:44:24 <libby> the vocab has 6 facets, each an rdf property

13:44:46 <libby> ...at the end of each one - there's a person, equiv to foaf:Person

13:45:28 <libby> ...journalists' viewpoints on things - breakign it doen to the simple journalistic questions

13:45:57 <libby> ...user interface - a heirarchy - fits with xml

13:46:24 <libby> ...users want features not standards - e.g. using outline paradign as front, can have rdf behind

13:46:35 <libby> ...layer cake - all the end users sees is cake!

13:46:55 <libby> everuone laughs at: http://semtext.org/2004-02/slides/img4.html :)

13:47:03 <libby> ...homer is a good end-user

13:47:08 <libby> [live demo]

13:48:22 <libby> ...uses a css class and javascript to change rendering

13:50:10 <libby> ...in the xhtml dendering, you can put it through xslt to get an rdf version

13:50:22 <libby> rendlering

13:50:32 <libby> ...the icons are tranformed to properties

13:51:16 <libby> ...compare a picture of the rdf to the end user

13:51:31 <libby> danbri: are you representing the tree state as RDF?

13:51:47 <libby> dannyayers: can be done

13:52:46 <libby> ....foaf is making good ground getting to end users - viva foaf

13:52:53 <gk> gk is now known as gk-scr

13:53:07 <gk-scr> Libby Miller presents on image annotation.....

13:53:36 <gk-scr> (setting up)

13:54:02 <yonderb0y> yonderb0y is now known as yonderboy

13:54:09 <gk-scr> (no URI yet)

13:54:18 <gk-scr> "Some RDF image work"

13:54:45 <gk-scr> co-depiction... people in same image

13:54:52 <danbri> [kendall, do you want to follow libby's talk directly after?]

13:54:56 <gk-scr> exif data (digital cameras)

13:55:13 <gk-scr> annotating W3 conference images

13:55:15 <kclark> danbri: that might make sense, yes

13:55:33 <danbri> [then maybe break for coffee, leaving 4 talks for remains of afternoon. i might do one too if room]

13:55:43 <danbri> [ok u r next]

13:55:45 <gk-scr> Jim Ley and others, means to describe depiction in /region/ of an image

13:55:54 <kclark> danbri: if i don't cough to death

13:55:54 * danbri solicits a post-coffee scribe

13:56:08 <gk-scr> (picture of three plastic ducks)

13:56:33 <gk-scr> user interface allows marquee selection, and then add annotation in form

13:56:39 * danbri realises he can use irc timestamps for timing speakers

13:56:48 * bijan notes that mindswap has such a tool as well

13:56:49 <gk-scr> (That was Masahide Kanzaki work)

13:56:55 <kclark> bijan: i'm demoing it next

13:57:09 * bijan goes back to lurking

13:57:18 <gk-scr> next, annotating an image with stories associated

13:57:20 <kclark> danbri: i had a semchimp plugin that would time speakers or irc events; basically, tell me when foo seconds has passed

13:57:53 <kclark> hmm, actually, libby's doing so well i won't have much to say :>

13:57:53 <gk-scr> mouse over image highlights region and shows anotation

13:58:07 <danbri> [cool; ralph's zakim bot does that too, but requires an audio conf to be running i think. also has 'pick a scribe' thingie :]

13:59:08 <gk-scr> "Referential integrity" [?]... being able to refer to same person .... finding all instances of same person (?)

13:59:39 <gk-scr> (scribe thinks trying to describe demonstration of images is losing game ;-)

14:00:51 <gk-scr> Key thing is allowing search of separate data source to allow short identifying strings; content-based search.

14:02:02 <gk-scr> Vocabularies pertaining... foaf, exif, Jim Ley's SVG work, Jennifer Golbeck, creative commons, dublin core, RDFiCal, Geo...

14:02:12 <gk-scr> Mainly,\mixing is important.

14:02:27 <gk-scr> Trick to work out how things shoukld link together

14:03:23 <gk-scr> W3 Photo project. Lots of issues need writing up properly.

14:04:21 <gk-scr> distributed - many creation tools, redistribution of data, single liberal licence, identification of people, user interface.

14:04:50 <gk-scr> DanW: is "referential integrity" a "term of art"?

14:05:13 <gk-scr> Libby: it has special meaning for hypertext, (missed rest of answer)

14:05:32 <gk-scr> --------

14:05:46 <gk-scr> End session, back at 3:35

14:05:50 <gk-scr> --------

14:06:05 <gk-scr> gk-scr is now known as gk

14:13:45 <Liddy> hi

14:14:49 <D[a]vey> D[a]vey is now known as Davey

14:18:11 <tav> tav is now known as tav|offline

14:34:21 <tav|offline> tav|offline is now known as tav

14:38:44 <chaalsNCE> anyone here know how to run chump?

14:39:04 <chaalsNCE> There is an instruction to "quote # characters" - any idea how?

14:39:07 <yonderboy> you mean install or use?

14:39:22 <edd> just prefix the # on the unix commandline with a \

14:39:29 <chaalsNCE> Thanks...

14:39:37 <edd> or put "" around the whole channel name, e.g. "#rdfig"

14:40:25 * dajobe $world =~ s/rdfig/swig/

14:40:53 <edd> maybe we should put in an alias for swig.xmlhack.com or something

14:41:18 <kao> dajobe: does this include the mailing list and this channel?

14:41:41 <dajobe> not for me to do that

14:41:56 <dajobe> I just changed the chatlogs & chump

14:43:06 * edd just made swig.xmlhack.com point to the right place

14:43:13 <edd> well, in a second

14:43:31 <libby> ----kendal clark

14:43:37 <libby> People and Images

14:43:43 * danbri thanks libby for scribing

14:43:56 <libby> ...client is US defence community

14:44:23 <libby> 'photostuff' - java sclient, grad student project

14:44:42 <libby> ...difference - can load any owl ontoollgy and use it to make assertions about arbitrary images

14:45:26 <libby> ...communication with remote services for user interface - what we have called 'referential integrity' - esp with many different spellings of 'osama bin laden'

14:45:27 * chaalsNCE uunderstaands

14:45:37 <libby> ...instance support

14:45:45 <libby> ...SVG outlines to represent image regions

14:45:50 * danbri was thinking of chaaaals at that point :)

14:46:02 <libby> ...building a better version now - new suite of tools for image annotation++

14:46:25 * chaalsNCE understands the problems of finding mr M[a]c[C]a[r]th{ie,y}Nevil[l][e]

14:46:36 <libby> ...would like to do some inferencing beyond codepiction - codepiction is less important for intelligence community

14:47:06 <libby> ...bad guy 1 at event A - anotrher image saying he was elsewhere at same time - computation geography to prove that they can;t be the same person

14:47:21 <libby> [demo - shows danny and norm]

14:47:46 <libby> (http://swordfish.rdfweb.org/photos/2004/03/01/2004-03-01-Pages/Image18.html) in fact

14:48:18 <libby> ...can load ontologies off the web e.g terrorism ontology ('terrorists are people too' ;)

14:48:28 <libby> ...arbittary outline

14:48:41 <libby> [danny erk's as he is labeled as a terrorist ;)]

14:49:06 <libby> ...drag and drop interface for types

14:49:16 <libby> ...a form driven by the ontology

14:49:29 <libby> ...instance support - other instances of class 'terrorist-leader'

14:49:55 <libby> ...can submit via http to the site

14:50:09 <libby> swint.mindswap.org

14:50:23 * chaalsNCE wonders if we can have concentric target regions....

14:50:27 <libby> ...please have a play with this, suggestions for improvmenet welcome

14:50:32 <libby> ---ends

14:50:53 * edd cheers kendall

14:51:11 <libby>http://swint.mindswap.org/

14:51:11 <dc_rdfig> N: http://swint.mindswap.org/ from libby

14:51:22 <libby> N:|Experimental SWINT web page

14:51:22 <dc_rdfig> Titled item N.

14:51:34 <libby> N:"consider all data as fake"!

14:51:34 <dc_rdfig> Added comment N1.

14:52:35 <libby> ---Danny Weitzner

14:52:46 <libby> "Transparaency Paradox"

14:52:49 <RalphS>http://www.w3.org/2004/Talks/0302-trans-privacy-swig/

14:52:49 <dc_rdfig> O: http://www.w3.org/2004/Talks/0302-trans-privacy-swig/ from RalphS

14:52:52 <dmiles> dmiles is now known as dmiles_afk

14:53:09 <RalphS> O:| The Transparency Paradox -- D.J. Weitzner

14:53:09 <dc_rdfig> Titled item O.

14:53:13 <libby> - kerchoffs' principle - security by obscurity doesn;t work

14:53:14 <dmiles_afk> dmiles_afk is now known as dmiles

14:53:16 <dmiles> bijan: have you ever used java?

14:53:20 <libby> thanks RalphS

14:53:30 <dmiles> (erm i mean coded in it)

14:53:30 <kclark> nmg: http://owl.mindswap.org/

14:53:31 <libby> ...2 examples

14:53:37 <kclark> N:http://owl.mindswap.org/

14:53:37 <dc_rdfig> Added comment N2.

14:53:45 <libby> ...1. TIA - total information awareness

14:53:58 <kclark> N:http://mindswap.org/2003/PhotoStuff

14:53:58 <dc_rdfig> Added comment N3.

14:54:29 <libby> ...after sept 11, pentagon decided needed to get better at tracking terrorists

14:54:48 <libby> ...terrorist behaviour leave signatures in databases - e.g. flights

14:55:38 <libby> ...similar to tracking submarines through their signatures

14:55:44 <danbri> [chaals, u r nxt]

14:56:00 <kclark> danbri: sorry about turning foaf to evil :<

14:56:00 * chaalsNCE would be more re-assured if they *were* needing new powers of information collection...

14:56:13 <libby> ...givet regularly looks in these dbs but now can learn more. outcry - but noone should be worried - they were lookign at info already auth to look at

14:56:14 <danbri> :)

14:56:51 <libby> ...dillemma for us - the legal tools we have in anglo-european coupltries focus in privacy protection by limiting the coolection of information in one way or another

14:57:02 * danbri turned the second hand back on his laptop clock display, can actually keep time now

14:57:15 <libby> ...suggests that so much more info is now available in public or quasi-public places, privacy protection through collecvtion l;imittaion is not adequate

14:57:26 <libby> ...esp because of power of inferencing

14:57:40 <sh1m|cannes> sh1m|cannes is now known as sh1m

14:57:43 <libby> ...one answer - shared private computatio n- doesn;t think those work

14:57:52 <sh1m> sh1m is now known as sh1m|cannes

14:58:00 <libby> ....privacy intrusion increased geometrically with the number of triples on the sw ? - thinks so

14:58:19 <libby> ...thinks shouel embrace tranparency, show people how much inmform,ation is avilable about them

14:58:22 <libby> ...how exposed they are

14:58:35 <libby> ...legal matter - use limitatios rather than collection limitiation

14:59:11 <libby> ...apologies to David brin..but embracing tranparaecy requires good dialogue between policy makers tech people and?? community

14:59:24 <Liddy> ...Charles McN....

15:00:17 <Liddy> thesaurus work and RDF for people with disabilities

15:00:27 <DannyNCE> Danny's talk: http://www.w3.org/2004/Talks/0302-trans-privacy-swig/

15:00:28 <Liddy> ...people you meet ...

15:01:17 <Liddy> meatdata activity of a typical kind...and a thesaurus...and software to match and merge concepts across thesauri..

15:01:23 <danbri> see also http://lists.w3.org/Archives/Public/public-esw-thes/ list

15:01:42 <danbri> (recent threads from folk in library community)

15:02:00 <Liddy> ..many people use symbols instead of text for a variety of reasons incl disabilities

15:02:05 <kclark> N:Forgot to mention that Photostuff also supports annotations of video segments (buggily, though)

15:02:05 <dc_rdfig> Added comment N4.

15:02:18 <Liddy> ..simple grammars and limited vocabs often for concrete things

15:02:38 <Liddy> a bit like sign langauges..

15:02:41 <danbri> [bwm, u r on next; ok?]

15:03:08 <Liddy> ..variety of symbol sets...copyright issues...

15:03:28 <Liddy> ...photos can be useful

15:04:24 <Liddy> ..sometimes the symbol is so concrete that eg a person misses a bus becuase it is green and there was a red one on their instructions

15:04:54 <Liddy> WWAAC are mapping across symbol languages...

15:05:39 <Liddy> ...eg a picture of a cup might be p[laced in text to indicate a cup ..

15:05:58 <Liddy> ...end...

15:06:23 <Liddy> ps - the poeople doing this work are now using the thesaurus ideas...

15:06:39 <libby> liddy are you scribing?

15:06:44 <Liddy> this time yep

15:06:46 <libby> ----brian mcBride, HP labs

15:06:48 <libby> cool

15:06:54 <Liddy> ..Brian McBride..

15:07:17 <Liddy> HP masters asking about why Sem wrk is being done..

15:07:40 <Liddy> ..to do this, best to show a spec of running code and ask if it is interesting...

15:08:28 <Liddy> ..this is an eg: (he has screen problems)

15:09:05 <Liddy> ..the target HP people wanted a mobile phone world ..that could have bits chained together...

15:09:16 <Liddy> eg list of movies, reviews, taxis, etc

15:09:36 <chaalsNCE> My example is actually the sort of content they generate. By picking symbols, you collect the associated RDF, which can be used to put the right kind of symbol, or text, or whatever.

15:09:41 <kclark>http://monkeyfist.com/kendall/pstuff.pdf

15:09:41 <dc_rdfig> P: http://monkeyfist.com/kendall/pstuff.pdf from kclark

15:09:58 <kclark> P:|Photostuff: Image Annotation Presentation

15:09:58 <dc_rdfig> Titled item P.

15:10:03 <Liddy> ...words have 'semantic links' or queries that go to services related to the original thing eg a movie

15:10:06 <kclark> peepo: (PDF)

15:10:09 <chaalsNCE> ... but they didn't have SVG so were doing it in a big complex java tool, rather than the kind of thing amenable to javascript hackery, as a first-stage development)

15:10:33 <Liddy> ..on a phone you want a very smll link - not to type uris etc..

15:10:49 <danbri> [barcode readers, rfids etc, bluetooth also useful here for avoiding data entry, imho]

15:11:23 <Liddy> ..it also knows where you are so it picks up cinema info

15:11:35 <Liddy> ..which leads to transport services

15:12:01 <kclark> rfid++ (well, rfid--, but what can you do?)

15:12:16 <danbri> [yep]

15:12:23 <danbri> [i want some rfid kit]

15:12:31 <Liddy> ..telcos like this because it provides a way of selling things

15:12:39 <kclark> P:(PDF)

15:12:39 <dc_rdfig> Added comment P1.

15:12:47 <libby> - people who like taxis also like busses :)

15:12:54 <kclark> danbri: niel b. and i are writing about it for o'reilly

15:13:05 <dirkx> danbri: www.hidcorp.com -> serial port version; excelent to integrate.

15:13:10 * chaalsNCE notes that might not be the case for taxi company owners

15:13:16 <Liddy> ..Danbri - local sensitive services are often seen to be the way but are the telcos standing in the way? when might it really happen?

15:13:51 <Liddy> BmcB - there is money related to telco use of the web unlike on the other internet web

15:14:03 <Liddy> ...end

15:14:04 <dirkx> Next preso on line at http://www.asemantics.com/presos/tw-2004-cannes/ - rough and live strawman demo pages on http://foaf-demo.asemantics.com/ex.html

15:14:30 <Liddy> ...Dirkx...

15:15:46 <Liddy> ..you want stuff about the page not the page itself..

15:16:37 <Liddy> ..benefits of DNS..

15:17:16 <areggiori> or you want to point to something that does not have a URI e.g. foaf:Person, ical:Event

15:17:22 <Liddy> esp DNS anticipating what you want to know..

15:17:43 <Liddy> ..NAPTR-DNS ..

15:18:18 <danbri> see http://ibm.com/developerworks/oss/lsid/

15:18:43 <dajobe>http://www.asemantics.com/presos/tw-2004-cannes/

15:18:43 <dc_rdfig> Q: http://www.asemantics.com/presos/tw-2004-cannes/ from dajobe

15:18:50 <libby>http://www-124.ibm.com/developerworks/oss/lsid/

15:18:50 <dc_rdfig> R: http://www-124.ibm.com/developerworks/oss/lsid/ from libby

15:18:52 <dajobe> Q:|Tell me about that URI, Dirk

15:18:53 <dc_rdfig> Titled item Q.

15:19:14 <areggiori>http://www.i3c.org/wgr/ta/resources/lsid/docs/

15:19:14 <dc_rdfig> S: http://www.i3c.org/wgr/ta/resources/lsid/docs/ from areggiori

15:19:16 <libby> R:|LSID resolution Propptocol project

15:19:16 <dc_rdfig> Titled item R.

15:19:44 <Liddy> using DDDS rfc3401-5

15:19:52 <areggiori> R:|LSID protocol OMG standard/proposal to I3C

15:19:53 <dc_rdfig> Titled item R.

15:20:19 <libby> ta

15:21:09 <Liddy> ..all sorts of terminal results..

15:21:09 <areggiori> R:which covers a more general solution to the problem and kinda "registration" protocol for URI-to-URI mapping and its metadata

15:21:10 <dc_rdfig> Added comment R1.

15:21:54 <Liddy> ..benefits of technique..

15:23:21 <areggiori> Q:advantage of DNS/DDDS solution for this kind of "mapping" is that it builds on existing DNS deployed technology and infrastructure

15:23:22 <dc_rdfig> Added comment Q1.

15:23:26 <danbri> testing.... [[ dig dna:~ danbri$ dig NAPTR www.asemantics.com ->

15:23:28 <danbri> ;; ANSWER SECTION:

15:23:28 <danbri> www.asemantics.com. 1800 IN NAPTR 100 10 "u" "http+I2C" "!^http://([^:/?#]*)(.*)$!http://\\1/rdf.pl/\\2!i" .

15:23:28 <danbri> www.asemantics.com. 1800 IN NAPTR 100 10 "u" "http+I2I" "!^http://([^:/?#]*)(.*)$!http://\\1/url.pl/\\2!i" .

15:23:28 <danbri> www.asemantics.com. 1800 IN NAPTR 100 10 "u" "http+I2N" "!^http://([^:/?#]*)(.*)$!http://\\1/urn.pl/\\2!i" .

15:23:30 <danbri> www.asemantics.com. 1800 IN NAPTR 100 10 "u" "http+I2R" "!^(.*)$!\\1!i" .

15:23:32 <danbri> ]]

15:23:45 <danbri> (this might not be quite what he's talking about in his example, i miss subtleties]

15:24:30 <Liddy> danbri - what are the downsides..

15:24:47 <areggiori>http://foaf-demo.asemantics.com/ex.html

15:24:47 <dc_rdfig> T: http://foaf-demo.asemantics.com/ex.html from areggiori

15:25:24 <areggiori> T:page with examples and explaination of dirk tried to outline/explain

15:25:24 <dc_rdfig> Added comment T1.

15:25:25 <Liddy> ..if there is not the info that's needed, nothing comes abck, of course..

15:25:46 <areggiori> T:|DDDS for FOAF - examples/demo

15:25:46 <dc_rdfig> Titled item T.

15:26:43 <Liddy> Danbri - a quick 1 minute soapbox for everyone.

15:27:09 <Liddy> pls say name clearly before you start!!

15:27:30 <libby> Jos deRoo

15:27:50 <libby> - bottom up dataaccess working group and BP wg - happy to contribute

15:27:53 <Liddy> Jos De ros - after a long time ...

15:28:04 <dirkx> downsides: extra step in the path; extra level of indirection through a medium (DNS) which is alien to some webby people.

15:28:05 <danbri> Liddy vs Libby for scribing? Liddy you ok doing this?

15:28:08 <Liddy> real world problem solving is good and increases confidence...

15:28:35 <Liddy> name?

15:28:52 <areggiori> Q:downsides - extra step in the path; extra level of indirection through a medium (DNS) which is alien to some webby people.

15:28:53 <dc_rdfig> Added comment Q2.

15:29:00 <danbri> -> http://www.swi.psy.uva.nl/usr/Schreiber/home.html Guus Schreiber

15:29:06 <Liddy> using RDF interest group experiences in best practices world..

15:29:09 <chaalsNCE> Guus Schreiber

15:29:11 <Liddy> image work interesting

15:29:37 <danbri> -> Dave Beckett

15:29:51 <danbri> -> Chris Bizer

15:29:58 <Liddy> Dave Beckett has already spoken

15:29:58 <danbri> dajobe, no comment?

15:30:06 * chaalsNCE will scribe liddy

15:30:07 <Liddy> Chris too..

15:30:27 <danbri> ppl like best practices stuff

15:30:38 <chaalsNCE> Liddy: THink it is exciting to hear real world practices being organised - will be helpful for people who are a couple of steps away from teh RDF world

15:30:42 <danbri> -> Alan Tutor(?), ...Foundation, spain

15:30:47 <jose-tp> Alain Chuter

15:30:51 <chaalsNCE> Alan Chuter, Fundacion ONCE

15:30:53 <danbri> ...interesting in nav aids for disabled users

15:30:55 <jose-tp> Fundacion Once

15:31:00 <danbri> -> Brian McBride HP

15:31:08 <danbri> ...important: next thing is deployment

15:31:12 <chaalsNCE> most important thing is deployment. The interest is there

15:31:22 <libby>http://www.ilrt.bris.ac.uk/discovery/2004/03/swig-images/

15:31:23 <dc_rdfig> U: http://www.ilrt.bris.ac.uk/discovery/2004/03/swig-images/ from libby

15:31:24 <danbri> ...have enough interest from academic/research that i'm confident underlying tech base will keep maturing

15:31:24 <chaalsNCE> the tech base will develop cos it has momentum

15:31:31 <chaalsNCE> important to build bridges with XML community

15:31:36 <danbri> important to build bridges w/ xml community

15:31:38 <libby> U:|Libby's images presentation (mostly links)

15:31:38 <dc_rdfig> Titled item U.

15:31:54 <danbri> ...must end isolation of the community; shout about cool stuff u build.

15:32:09 * edd is of half a mind that the xml community is dispersing

15:32:18 <danbri> Charles McCathieNevile -- I'd like to see the cool things ppl have been building explained clearly

15:32:47 <danbri> ...ie not frustrating ppl by using abbreviations that exclude large % of audience for sake of a few words

15:32:51 <chaalsNCE> Lisa Seeman, UBaccess

15:32:52 <danbri> -> Lisa Seeman

15:32:59 <danbri> ...we're using rdf for accessibility techniques

15:33:05 <chaalsNCE> using RDF to solve accessibility problems, WCAG compliant sites, ...

15:33:16 * danbri hands stolen scribe baton to charles

15:33:21 <chaalsNCE> (don't say WCAG, say Web Content Accessibility Guidelines)

15:33:36 <chaalsNCE> the power of RDF is really cool for solving problems.

15:34:07 <chaalsNCE> anyone who is interested can come to a meeting on thursday for the Web Content Accessibility Guidelines to look at techniques...

15:34:18 * bijan wonders what power, which problems

15:34:36 * chaalsNCE would love to have a chat with Bijan about it.

15:34:51 * chaalsNCE suspects Lisa would like Bijan to buy SWAP as a way of seeing cool things that can be done...

15:34:59 <chaalsNCE> Chuck Myers, Adobe

15:35:16 <chaalsNCE> We need to look at what people are trying to do and how to explain solutions to their problems in their language

15:35:26 <chaalsNCE> Patrick Stickler, Nokia

15:35:32 <chaalsNCE> Pleased to see the Recommendations.

15:35:42 <chaalsNCE> Patrick Stickler agrees with Brian McBride

15:35:43 <chaalsNCE> Patrick Stickler agrees with Brian McBride

15:35:45 <chaalsNCE> Patrick Stickler agrees with Brian McBride

15:35:49 <chaalsNCE> Foundations are in place

15:35:58 <chaalsNCE> to build applications.

15:36:47 <bwm> patrick thinks bootstrapping knowledge is important and overlooked

15:37:00 <chaalsNCE> Alberto Reggiori, @semantics

15:37:03 <bwm> alberto - sw is now moving

15:37:17 <chaalsNCE> we see things moving, and we are starting to build applications on it so we can try and sell something to people

15:37:42 <chaalsNCE> we are missing things that need to be worked on, but we like where things are going. And like the Best Practices stuff

15:37:53 <chaalsNCE> Dirk Willem Van Gulijk, @semantics

15:38:02 <chaalsNCE> standards stuff are there which is really important

15:38:21 <chaalsNCE> even at the bottom layer you can already do cool stuff - don't wait for everything to be done perfectly before you start doing work.

15:38:33 <chaalsNCE> There are bits missing - the URI problem is part of a larger problem.

15:39:01 <chaalsNCE> namies/locations stuff.... there is technology that exists and you can do it without waiting for it to be cast in stone.

15:39:08 <chaalsNCE> Jean ?? INRIA

15:39:25 <chaalsNCE> posted a bibliography on conceptua graphs and RDF

15:39:33 * chaalsNCE missed name

15:39:39 <chaalsNCE> ?? INRIA

15:39:58 <chaalsNCE> pleased with discussion. Liked the lightning talk with the cup in it (chaals' example)

15:40:09 <chaalsNCE> Too much semantics and not enough semiotics is common finding

15:40:21 <chaalsNCE> even if the answer is great you need to know how to show it to the user

15:40:37 <chaalsNCE> would like to keep working on semantics and semiotics and how to show them in different contexts

15:40:45 <chaalsNCE> Reto Bachmann-Guer

15:41:00 <chaalsNCE> Was pleased to meet people whose name I know from email - that is very valuable.

15:41:15 <danbri> [good point; huge email burden easier to bear if you can put names/faces/personalities to mail msgs]

15:41:25 <chaalsNCE> The input of ideas on how to use RDF and how other people are using it was great - Yoshio's stuff about probabilistic reasoning for example

15:41:38 * bijan 's 1 minute shout out: HELP! I'M IN YET ANOTHER WEB SERVICES WORKING GROUP!!

15:41:43 <bijan> :)

15:41:49 <libby> poor bijan

15:41:55 <bijan> C'est ma vie

15:42:00 <chaalsNCE> I am frightened about overloading with new ideas. I think we need to learn how to deal with ambiguity, not have an academie francaise (RAE is the spanish equivalent)

15:42:06 <chaalsNCE> Tom Croucher, Uni Sunderland

15:42:14 <chaalsNCE> RDF is a great way to give hooks to processing information.

15:42:40 <chaalsNCE> the amount of info for machines currently is very limited. We will be able to cope with the deluge of info if we can pass it off to machines.

15:42:48 <chaalsNCE> I am turning to using RDF in digital libraries

15:42:55 <chaalsNCE> CMN remembers he is intersted in work in other languages

15:43:00 <chaalsNCE> Phil Tetlo, IBM

15:43:24 <chaalsNCE> Vision of ....

15:43:30 <danbri> [man fights laptop]

15:43:41 <chaalsNCE> [Brian McBride falls into the deathtrap and drags mummy's machine with it]

15:43:47 <chaalsNCE> Ralph Swick W3C

15:44:00 <chaalsNCE> This forum has proved valuable - mixture of different types of users developers and so on.

15:44:31 <chaalsNCE> Would encourage you to remember the learning curve you went through, and document it so that the rest of the people get some signposts as they start down the path we are travelling on

15:44:43 * chaalsNCE apologises to Phil for missing his comments

15:45:09 <chaalsNCE> Andy Seaborne HP

15:45:29 <chaalsNCE> Would encourage people to publish vocabularies and code - seeing what people use and is useful will help people to learn by examples

15:45:34 <chaalsNCE> that they can add to

15:45:44 <chaalsNCE> Yoshio, Panasonic / W3C

15:45:52 <chaalsNCE> Glad to see technologies coming to flower

15:46:03 <chaalsNCE> Lots to do still. Lots to think about

15:46:11 <danbri>http://www.w3.org/TR/NOTE-MCF-XML-970624/simple.gif

15:46:17 <dc_rdfig> V: http://www.w3.org/TR/NOTE-MCF-XML-970624/simple.gif from danbri

15:46:27 <danbri> V:|The image that taught me how MCF works

15:46:28 <dc_rdfig> Titled item V.

15:46:31 <chaalsNCE> Should be careful that we are giving more and more power to the Web - maybe we give it too much power

15:46:37 <danbri> V:And readied my brain for the RDF spec :)

15:46:39 <dc_rdfig> Added comment V1.

15:46:42 * chaalsNCE wonders if somone has chumped the Matrix recently (or Tron)

15:47:07 <chaalsNCE> We should keep people aware that what comes from the Web is based on a lot of people's assumptions. Trust and authentication technology is more an more important

15:47:11 <chaalsNCE> Libby Miller, ILRT

15:47:24 <sh1m|cannes> (Phil Tetlow says) My vision is a self-organizing web demostrating purpose through the formal use of semantics!! It has been a real pleasure listening to the gra

15:47:43 <sh1m|cannes> great work of the group ... Many thanks

15:47:48 <chaalsNCE> Excellent meeting. Best Practices is cool. I am going to investigate ways of looking at data we have to help people catalog their information, giving people interefaces that hep them

15:47:58 <chaalsNCE> interested in Danny Weitzner's privacy presentation / ideas

15:48:01 <chaalsNCE> Danny Ayers

15:48:09 <chaalsNCE> Grateful for the openness of this.

15:48:25 <chaalsNCE> Seems like the Semantic Web is starting to happen.

15:48:34 <chaalsNCE> There is a need to get the deployment and data moving.

15:49:05 <jose-tp> (is there any repository that repertoriates existing RDF applications, demos; not just ontologies?)

15:49:09 <yonderboy> S:|I3C: Life Sciences Identifiers

15:49:09 <chaalsNCE> One aspect that needs more attention is looking at XML applications that are bending over backwards to do something that is easy in RDF. Raising awareness by deploying things ...

15:49:09 <dc_rdfig> Titled item S.

15:49:12 <areggiori> I fully agree with Danny

15:49:18 <chaalsNCE> Graham Klyne, nineBynine

15:49:33 <chaalsNCE> Have been buried in technology. Top 3 goals are now applications (3 times)

15:49:43 <chaalsNCE> Looking for real world problems that existing tools can solve

15:49:58 <chaalsNCE> Also want to look at how to feed tools we have from existing data

15:50:03 <chaalsNCE> jeremy Carroll, HP

15:50:19 <chaalsNCE> Deployment and applications are important. But I'm not personally interested.

15:50:35 <chaalsNCE> I'll try to work on fixing gaps and cleaning up things that aren't yet finished.

15:50:47 <chaalsNCE> Maybe RDF 2 - after we have made som money

15:50:58 * chaalsNCE says BUY more Printer Cartridges!

15:51:07 <chaalsNCE> dan brickley, W3C

15:51:19 <chaalsNCE> There's masses of stuff at W3C and its all great. And we'll make more.

15:51:26 <chaalsNCE> There is lots of intellectual headd-hunting to do

15:51:38 <chaalsNCE> find people and get them owrking on the semantic web.

15:51:54 <chaalsNCE> The tech work won't stop, and I don't feel a big urge for standards tracking stuff.

15:52:05 <chaalsNCE> Big issue is convergence/aggregation/integration

15:52:28 <chaalsNCE> time to invest effort in puting these specifications together and showing how they work well together.

15:53:03 <chaalsNCE> Sharing success stories and how-tos - how lots of people got RDF and SVG work well together for lots of people without lots of effort.

15:53:09 <chaalsNCE> Over the next year or so I will be

15:53:24 <chaalsNCE> 1. Chasing funding to keep employed (scribe has great sympathy :-)

15:53:44 <chaalsNCE> 2. Building stuff. I am drawn to social, legal cultural political impact

15:54:01 * chaalsNCE thinks this is no surprise from teh grandfather-of-a-terrorist-vocabulary :-)

15:54:12 <chaalsNCE> Thanks everyone for coming....

15:54:20 <libby> A:[http://swordfish.rdfweb.org/discovery/2001/08/codepict/aboutevent.jsp?uri=http://www.w3.org/2001/sw/interest/meetings/tp2004|a few more photos from the meet] apologies to Danny Ayers for various photo-related embarrassments througout the day...

15:54:20 <dc_rdfig> Added comment A2.

15:55:17 <dajobe> dajobe has changed the topic to: Semantic Web Interest Group - weblog http://rdfig.xmlhack.com/ and chatlogs http://www.ilrt.bris.ac.uk/discovery/chatlogs/rdfig/

15:56:55 <chaalsNCE> It was Jean-François Baget who published bibliography (see above...)

16:02:11 <jeen> anybody know of an easy to 'filter' an OWL ontology and end up with an RDF Schema? Any XSL stylesheets for that purpose?

16:03:03 <bijan> Er..

16:03:10 <bijan> What properties do you want to preserve?

16:03:14 <bijan> Any?

16:03:19 <chaalsNCE> Good news - Brian and me mummy's computer are both basically fine

16:03:57 <jeen> I'd like to just map DatatypeProperty etc. to rdf:Property, and owl:Class to rdfs:Class, that sort of thing.

16:04:01 <libby> phew chaals

16:04:12 <bijan> Ok, just recongize you might be seriously distoriting the meaning :)

16:04:28 <bijan> I don't know of such a tool, but with any reasonable rdf api, shouldn't be hard at all

16:05:43 <jeen> well you might lose data, but I think it should be possible to avoid adding 'wrong' facts.

16:06:08 <jeen> alright well, i can hack something myself, I just hoped someone else had already done it :)

16:06:12 <bijan> Sorry :)

16:10:01 <dirkx> If your network seems dead; is generally just dns. Just sudo the line "nameserver 127.0.0.1" in "/etc/resolv.conf" and sudo named (on macosx).

16:17:14 <swh> swh is now known as swh_coffee

16:37:45 <earle> yoiks, I missed everyone.

16:39:10 <LotR> silly conference people moving outside of their 802.11 range ;)

16:42:42 <earle> Heh.

16:47:03 <danbri> thanks everyone :)

16:47:07 * danbri shuts down

17:00:18 <xkota> join foaf

17:00:28 <xkota> oops, excuse me.

17:06:23 <iwai> iwai is now known as iwaiNude

17:41:36 <swh_coffee> swh_coffee is now known as swh

17:53:42 <iwaiNude> iwaiNude is now known as iwaim

17:54:51 <bryce> bryce is now known as bryce_mmm_food

18:05:40 <earle> earle is now known as grault

19:02:18 <mdupont> mdupont is now known as md-afk

19:06:02 <bryce_mmm_food> bryce_mmm_food is now known as bryce

19:08:26 <chaalsNCE_> chaalsNCE_ is now known as chaalsNCE

20:04:28 <MrEIso> MrEIso is now known as Remosi

20:10:39 <timbl-tag> timbl-tag is now known as timbl

21:09:44 <SirKatz> SirKatz is now known as jordan

22:31:36 <golbeck>http://trust.mindswap.org/trustProject.shtml

22:31:37 <dc_rdfig> W: http://trust.mindswap.org/trustProject.shtml from golbeck

22:32:09 <golbeck> W:| With 8 Days left in the competition, the trust network is up to 987 members

22:32:10 <dc_rdfig> Titled item W.

22:32:42 <golbeck> W: Thanks to everyone who joined so far

22:32:42 <dc_rdfig> Added comment W1.

22:32:59 <golbeck> W: Check out the new statistics, inferencing, and visualization features

22:33:00 <dc_rdfig> Added comment W2.

22:33:07 <golbeck> W: Remember - the top contributor on the Top Ten list on March 10th gets a free Trust t-shirt!!!

22:33:07 <dc_rdfig> Added comment W3.

22:33:22 <golbeck> W: </salesPitch>

22:33:23 <dc_rdfig> Added comment W4.

22:33:27 <mortenf> hi jen

22:33:34 <golbeck> hi, mortenf

22:33:53 <mortenf> thanks for the key, will add link in a few minutes, must test a little more...

22:34:05 <golbeck> that's cool. thanks for setting it up!

22:34:15 <mortenf> np

22:40:21 <mortenf>http://www.wasab.dk/morten/2004/02/trust/

22:40:21 <dc_rdfig> X: http://www.wasab.dk/morten/2004/02/trust/ from mortenf

22:40:32 <mortenf> C:|Rate your friends!

22:40:32 <dc_rdfig> Titled item C.

22:41:00 <mortenf> A Trust and Reputation Project helper

22:41:06 <mortenf> X:A Trust and Reputation Project helper

22:41:06 <dc_rdfig> Added comment X1.

22:41:11 <mortenf> X:|Rate your friends!

22:41:11 <dc_rdfig> Titled item X.

22:41:14 <mortenf> C:

22:41:14 <dc_rdfig>http://www.w3.org/2001/sw/BestPractices/

22:41:15 <dc_rdfig> Rate your friends!

22:41:29 <mortenf> C:|SWBP

22:41:29 <dc_rdfig> Titled item C.

22:41:31 <mortenf> C:

22:41:31 <dc_rdfig>http://www.w3.org/2001/sw/BestPractices/

22:41:32 <dc_rdfig> SWBP

22:42:48 <mortenf> X:Supply your FOAF file, create trust statements for golbeck's project

22:42:49 <dc_rdfig> Added comment X2.

22:43:06 <mortenf> X:and possibly win a T-shirt!

22:43:07 <dc_rdfig> Added comment X3.

22:43:42 * bryce admires the nifty gradient on the input field at http://www.wasab.dk/morten/2004/02/trust/

22:43:57 <mortenf> it's a rip-off from golbeck!

22:48:13 <golbeck> it looks better on your page, though

22:48:23 <dirkx> mortenf: you sure you do not want to include a quick email token/file-based token for auth purposes; very tempting to just take other people their FoaF :-)

22:48:31 <dirkx> either a token you get email - or a token you are asked to put in your own foaf for validation would do effectively..

22:48:33 <mortenf> i changed it to a png, and enlarged it a bit

22:49:00 <mortenf> right, it's not secure, it's just help :)

22:49:49 <mortenf> but yeah, i'll do that for the next round...

22:50:20 <dirkx> Well - for your research you could say; I only use those trust values which people acutally put in their foaf files themselves.

22:50:31 <bryce> now would be a good time to buy that spamable email addresses CD I've been wanting to get. Then I'd be sure and win the t-shirt after submitting 10 million addresses.

22:50:40 <dirkx> In any case; I've no idea how to answer trust in such a way.

22:50:56 <mortenf> it's not easy, that's for sure.

22:51:01 <dirkx> I.e. what sort of trust are you asking for ?

22:51:18 <mortenf> follow the link...

22:51:19 <dirkx> trust with my Pin code, trust with my girlfriend, trust on technology, trust to be reciprocal kind with me

22:51:30 <golbeck> i've been saying how likely is it to trust that this person will send you an email worth looking at

22:51:37 <dirkx> :-) and even following that link :-) not too clear.

22:51:38 <bryce> dirkx, it is possible to rate trust on specific subjects, also.

22:51:42 <golbeck> a 10 you'd look at right away, and a 1 you'd look at never, maybe

22:51:55 <golbeck> but a lot of people (me included) are just using a more general notion

22:52:13 <bryce> see http://trust.mindswap.org/trustOnt.shtml for example subject-specific trust ratings.

22:52:20 <golbeck> ie i trust my spouse a lot, i kind of trust my boss, and i trust that nasty ex very little

22:52:33 <IsoosI> hmm

22:52:39 <IsoosI> but you usually trust people for different things

22:52:48 <golbeck> true

22:52:51 <IsoosI> I trust my plumber to plum, but I wouldn't let him anywhere near my computer

22:52:51 <dirkx> sure - but even when keyed to context I am finding that i am redefinging how I mean trust in each category. and what for.

22:53:15 <golbeck> you can include a subject with each trust rating

22:53:23 <golbeck> and give several ratings to each person based on different subjects

22:53:26 <dirkx> And there are cases where I trust as it is the only option; and cases where I trust - but yet have no reason to create the situation so trust is needed.

22:53:45 <golbeck> but for now, i'm ignoring that so that i can actually show something on a large network

22:53:51 <dirkx> Right - but somehow that is not quite consistent; i.e. the same trust+subject specifier means different thigns across the board.

22:54:07 * golbeck once again refers everyone to:

22:54:10 <dirkx> Perhaps it should rather be trust-case/context instead of subject.

22:54:23 <golbeck>http://citeseer.ist.psu.edu/marsh94formalising.html

22:54:23 <dc_rdfig> Y: http://citeseer.ist.psu.edu/marsh94formalising.html from golbeck

22:54:35 <golbeck> ug...how to i unchump that?

22:54:42 <golbeck> Formalizing trust as a computational concept

22:54:55 <golbeck> dissertation by stephen marsh

22:55:08 <bryce> submit a blank or other url for Y in order to replace it. or just title it and let it be.

22:55:15 <golbeck> i certainly have no desire to solve the issue of what it means for a human to trust another

22:55:36 <golbeck> which is where discussions of this project often devolve

22:55:43 <dirkx> Sure - but that is computational trust assertions - we are now asked something a bit different :-)

22:56:07 <golbeck> i leave it to the user to decide what trust means and how to rate it

22:56:49 <golbeck> there are no good answers to these questions of how trust is different in different contexts and such

22:57:06 <yonderboy> bryce, how do you submit a blank or other url for Y?

22:57:13 <yonderboy> for example

22:57:18 <golbeck> so i happily agree to and support trust ratings wrt subjects, but i don't want to do define it more than that

22:57:29 <golbeck> for fear of getting sucked into a philosophy discussion

22:57:39 <golbeck> Y:

22:57:40 <dc_rdfig>http://citeseer.ist.psu.edu/marsh94formalising.html

22:58:39 <golbeck> A"

22:58:49 <mortenf> Y:=http://example.com/

22:58:49 <dc_rdfig> Replaced URL of Y.

22:58:51 <golbeck> Y:"

22:58:52 <dc_rdfig> Added comment Y1.

22:59:09 <golbeck> Y:""

22:59:09 <dc_rdfig> Added comment Y2.

22:59:12 <golbeck> ug

22:59:25 <edd> Y1:""

22:59:25 <dc_rdfig> Deleted comment Y1.

22:59:25 <edd> Y1:""

22:59:25 <dc_rdfig> Deleted comment Y1.

22:59:26 <golbeck> Y1:""

22:59:27 <dc_rdfig> Comment Y1 not found.

22:59:31 <golbeck> Y2:""

22:59:31 <dc_rdfig> Comment Y2 not found.

22:59:31 <mortenf> heh

22:59:41 * golbeck kicks dc_rdfig

22:59:44 <yonderboy> mortenf, thanks

22:59:50 <golbeck> Y:=""

22:59:52 <dc_rdfig> Replacement must be a valid URL.

22:59:58 <golbeck> sigh

23:00:03 <yonderboy> Y:

23:00:03 <dc_rdfig>http://example.com/

23:00:05 <edd> golbeck: there's a proper way to express your fury with the bot :)

23:00:07 <yonderboy> done

23:00:09 <edd> dc_rdfig: foo

23:00:09 * dc_rdfig falls down a well.

23:00:14 <golbeck> NICE

23:00:16 <mortenf> oh, btw, edd: is there a reason for foafbot's keys not being on the key server?

23:00:18 <mortenf> ;)

23:00:28 <edd> mortenf: indubitably

23:19:27 <grault> grault is now known as earle