IRC log of swarch on 2003-03-07

Timestamps are in UTC.

00:00:27 [libby]
(http://www.ksl.stanford.edu/projects/dql/syntax.shtml is a syntax url i found recently)
00:01:15 [libby]
returns sets of bindings.
00:03:40 [libby]
want to be able to specify justification....
00:05:51 [libbyscri]
ian: impleemnting the entire thing (e.g. queries with cycles) is hard. some restritions not so bad
00:05:53 [danbri]
I made a front page for rdfq test case repository, http://www.w3.org/2003/03/rdfqr-tests/
00:06:16 [danbri]
...haven't any content (test cases) yet, but gathered pointers to all existing test stuff i can find
00:06:19 [danbri]
...surveys etc
00:07:22 [libbyscri]
ericP - could you do a quick one-pass squish-like pass?
00:07:33 [libbyscri]
benjamin - yes, even if the backend was complex
00:07:46 [libbyscri]
ericP: wouldnt lose perfortmance on simple queries
00:08:07 [libbyscri]
Mike: can assert premises
00:09:02 [libbyscri]
ericP: variable feature: may bind, must bind, whether the node is a vraiable. also - find all results or work for a while and then return.
00:09:37 [libbyscri]
ericP: on the reporting - which variables do you want to see.
00:10:05 [libbyscri]
ericP: = 3 areas so far. wold like input. would like to do a taxonomy of query charactristics
00:10:25 [libbyscri]
josD: the same query can have different proofs.
00:10:55 [libbyscri]
ericP: in algae, you'll get multiple rows - different reasons
00:11:23 [libbyscri]
[missed some]
00:12:05 [libbyscri]
josd thinks this is a whole different dimension
00:12:15 [libbyscri]
ericP: exhaustive search/ result grouping
00:12:38 [libbyscri]
josd: 2 dimensions
00:13:08 [libbyscri]
raphael: kaon project - QL for ontologies (one binding is RDF)
00:14:09 [libbyscri]
...different datamodels for rel dbs, RDF dbs, ontology stores.
00:14:58 [libbyscri]
...retrurns not tuples but classes and properties - things that correspond to the datamodel (of onbtology)
00:15:12 [libbyscri]
....modelled by a set of datalog ruless
00:16:10 [libbyscri]
...easy composition of orthogonal operators
00:16:43 [libbyscri]
....can rewrite e.g. to SQL where posisble
00:17:33 [libbyscri]
benjamin: can oit bind against uri or an arbitrary literal (R: yes) - sounds like ruleml
00:18:23 [libbyscri]
...has things like, SOME, INVERSEOF, ....
00:18:40 [danbri]
timecheck -- how much longer do we have? Does the meeting run until 7.30pm?
00:19:09 [libbyscri]
that's right yep
00:19:21 [libbyscri]
oh, only 10 mins more...
00:19:55 [libbyscri]
....and, or and not, where not is negation-as-failure
00:22:06 [libbyscri]
ian: can;t express cycles in non-distinguished variables.
00:23:49 [libbyscri]
Raphael: 'oneof' is expressed as a !
00:25:04 [danbri]
cf http://www.w3.org/TR/owl-ref/#EnumeratedDatatype (hmm, probably not the best ref for owl:oneOf)
00:25:37 [libbyscri]
....http://kaon.semanticweb.org - can download it.
00:25:56 [libbyscri]
ericP: example query now in http://www.w3.org/2001/11/13-RDF-Query-Rules/
00:27:20 [libbyscri]
we decide to continue till 8
00:27:35 [libbyscri]
said talking about ruleml
00:27:50 [libbyscri]
...close to 40 participating groups
00:30:03 [libbyscri]
...several ruleml engines and translators avilable, eg mandarax. some free
00:30:28 [danbri]
is this channel still being logged?
00:30:37 [libbyscri]
that's a good q
00:30:43 [libbyscri]
I can get it if not
00:30:56 [libbyscri]
rssagent is still there
00:33:01 [danbri]
rrsagent, help?
00:33:01 [danbri]
I'm logging. Sorry, nothing found for 'help'
00:33:05 [danbri]
ok cool
00:33:21 [libbyscri]
but on http://www.w3.org/2003/03/06-swarch-irc only got partial
00:33:31 [libbyscri]
oh wait, maybe it's 'tomorrow'?
00:34:01 [libbyscri]
coudl you check danbri? e.g. http://www.w3.org/2003/03/07-swarch-irc (forbidden)
00:35:27 [libbyscri]
...plan to submit usecaes to alberto's and andy's repository (http://rdfstore.sourceforge.net/2002/06/24/rdf-query/)
00:36:13 [libbyscri]
....rulebase - GEDCOM - family relationships
00:36:22 [libbyscri]
...created by Mike Dean
00:37:11 [libbyscri]
josd: how do those things cope with unique names assumption? very important, as we found in testcases
00:37:31 [libbyscri]
mikeD: one source of data, so implicit unque names assumption
00:38:00 [libbyscri]
josd: could do daml:differentFrom based on syntaxtic diffences?
00:38:16 [libbyscri]
mikeD: could do, but don't. should preobably ahve somethign more
00:40:12 [libbyscri]
thanks :)
00:40:24 [danbri]
world ACL
00:41:02 [libbyscri]
Said: workign on an ecommerce demo
00:41:58 [libbyscri]
...on website soon
00:42:27 [libbyscri]
ericP: could we use the rules for rule languages?
00:43:20 [libbyscri]
benjamin: yes you can use this [...] but often yopu want actions triggered, which is beyond the QL scope
00:43:35 [libbyscri]
Said: can use outside services.
00:45:33 [libbyscri]
[??] why doyou can it object orientated?
00:45:42 [libbyscri]
(harold will talk about it later)
00:46:06 [libbyscri]
danbri: this event-triggering rules seems very difefrent from timbl's rules...
00:46:26 [libbyscri]
danbri worried about scope - 'if and then'
00:47:21 [libbyscri]
said: didnt really start off as a rules language
00:47:59 [libbyscri]
benjamin: built-ins are very common in commercial rules sytems.
00:49:50 [libbyscri]
harold talks about object-orientated ruleml
00:50:28 [libbyscri]
....ruleml and rdfs overlap
00:50:41 [libbyscri]
...can use oo ruleml as an rdf ql and rles language
00:51:00 [danbri]
I just added a bunch more test-related links to http://www.w3.org/2003/03/rdfqr-tests/
00:52:23 [libbyscri]
...the subject and object are detrminted by their position
00:54:53 [libbyscri]
[sorry scribe missed if that was the positional one or the oo one]
00:55:24 [libbyscri]
...the two versions can be translated using xslt
00:56:35 [libbyscri]
...harold works us through an example
00:56:43 [libbyscri]
(url?)
00:57:38 [libbyscri]
...can express e.g. which page was accessed by person. in ruleml, queries are aq special case of rules, including only the body.
00:58:28 [libbyscri]
...there's an issue with bnodes
00:59:28 [libbyscri]
....need to give bnodes ids? [scribe missing soem of this, coudl be wrong]
01:00:21 [libbyscri]
...can do conjunctive queries: use 2 atoms
01:00:28 [libbyscri]
...now generalize to bnodes....
01:02:24 [libbyscri]
scribe fading... :(
01:02:33 [danbri]
imho ruleml includes a proposal for a new RDF syntax
01:03:23 [libbyscri]
....seesm to use generated nodeids
01:04:02 [libbyscri]
....model theory can bbuild on ruleml's rdf-xml integrating data model via flogic or triple
01:04:19 [libbyscri]
josd: a very diffreent datamodel to RDF.
01:04:31 [libbyscri]
harold: higher-order syntactic sugar
01:05:08 [libbyscri]
josd: why 'object-orientated'?
01:05:41 [libbyscri]
harold: as RDF is OO. also all the descriptions of objects clustered togetger
01:05:53 [libbyscri]
said: not the same as a definition of an OO language
01:05:59 [danbri]
timecheck!
01:06:04 [danbri]
we should be winding up...
01:06:07 [libbyscri]
benjamin: does not rely on positionality - has a name for the variable.
01:06:09 [libbyscri]
yep
01:06:14 [chaalsBOS]
'night folks
01:06:28 [libbyscri]
harold 'subject-oriented'
01:07:23 [libbyscri]
we are 10 minutes over time
01:07:34 [libbyscri]
banjamin says: 4 minutes!
01:07:47 [libbyscri]
he is outlining note draft on ruleml
01:08:05 [libbyscri]
....has also been discussed in the joint committee
01:08:22 [libbyscri]
....requiremeents, play nicely w the rest of SW, and also ws, xquery
01:08:43 [libbyscri]
...different tpes of rules
01:09:07 [chaalsBOS]
2 minutes gone
01:09:18 [libbyscri]
...those that derive new beliefs (like RDF query). also action rules - actiosn, get info. transfoemation can be vieweed as derrivation....
01:09:43 [libbyscri]
...some RDF qs return bindings, some graphs
01:10:21 [libbyscri]
...vraious chaacteristics of rules
01:10:37 [libbyscri]
...OO-ness (non-positional)
01:11:07 [libbyscri]
...lits of first-order expressiveness...
01:11:13 [libbyscri]
s/lits/lots
01:11:48 [libbyscri]
...in teh markup need to talk about things being derrived. this KB derrived this other stuff
01:12:36 [libbyscri]
...using several KVs
01:12:43 [libbyscri]
KBs even!
01:12:56 [libbyscri]
...complimentary doc on usecaese in teh joint committee
01:13:51 [libbyscri]
...questions?
01:14:09 [libbyscri]
benjamin proposes we head to bar, declaring victory
01:14:32 [libbyscri]
ericP: this would be good foddder - clarifying different parts of query
01:14:43 [libbyscri]
b: 3-4 weeks, public draft
01:14:51 [libbyscri]
...will post to RDF rules
01:15:29 [libbyscri]
---scribe declares victory. ajourned.....
02:06:10 [em-lap]
em-lap has joined #swarch
02:49:39 [las]
las has joined #swarch
03:55:53 [Tantek]
Tantek has joined #swarch
03:56:25 [Tantek]
Tantek has left #swarch
10:44:27 [DaveB]
DaveB has joined #swarch
11:39:15 [AndyS]
AndyS has joined #swarch
11:40:45 [AndyS]
AndyS has joined #swarch
11:51:35 [AndyS]
AndyS has joined #swarch
13:20:16 [em-lap]
em-lap has joined #swarch
13:20:28 [em-lap]
ack... no opps
13:20:41 [Zakim]
Zakim has joined #swarch
13:20:53 [em-lap]
em-lap has changed the topic to: semweb arch tech plen meeting - 2002-03-07
13:26:17 [em-lap]
ericP, you here?
13:47:23 [PStickler]
PStickler has joined #swarch
13:49:49 [jhendler_]
jhendler_ has joined #swarch
14:09:05 [jhendler_]
jhendler_ has joined #swarch
14:12:17 [danbri]
danbri has joined #swarch
14:12:38 [danbri]
(we got stuck in traffic)
14:14:46 [JosD___]
JosD___ has joined #swarch
14:15:02 [pfps]
pfps has joined #swarch
14:15:06 [danb_lap]
danb_lap has joined #swarch
14:15:57 [danb_lap]
uj
14:17:33 [danb_lap]
timbl: interested in scoping new work areas, how much time things would likely take, etc
14:17:48 [danb_lap]
...seems to me from way layers are developing, Query is next item ready for standardisation
14:18:07 [danb_lap]
...i read thru various bits and pieces
14:18:31 [danb_lap]
...has anyone read thru this? (not many hands go up)
14:18:42 [danb_lap]
...ben and harold fed back some possible changes
14:18:47 [DanC]
DanC has joined #swarch
14:19:01 [danb_lap]
...see http://www.w3.org/2001/sw/meetings/tech-200303/query
14:19:12 [danb_lap]
...b/g includes AndyS and Alberto's use case repository
14:19:15 [bwm]
bwm has joined #swarch
14:19:19 [danb_lap]
...can register examples
14:19:24 [GuusS]
GuusS has joined #swarch
14:19:30 [danb_lap]
...going over that, you get a pretty good idea of what folk are doing w/ rdf query
14:19:34 [danb_lap]
rrsagent, help?
14:19:34 [danb_lap]
I'm logging. Sorry, nothing found for 'help'
14:19:55 [danb_lap]
...interesting. Most syntaxes are non-xml
14:20:15 [danb_lap]
...half of them use the word 'SELECT', ie. SQL mind set, sometimes 'FROM', 'USING' etc
14:20:24 [libby]
libby has joined #swarch
14:20:28 [danb_lap]
...some have some punctation around a chunk of rdf, and an ipmlication sign
14:20:31 [danb_lap]
...fall into various groups
14:20:58 [danb_lap]
...seems that concrete syntaxes may have questions, but the abstract syntax there is much commonality
14:21:05 [danb_lap]
...simplest might be versa, which is just a path
14:21:15 [danb_lap]
...this is a subset of a general rdf graph matching template
14:21:48 [danb_lap]
...if you look at the graph match, ie. rdf with holes, a question of what you can put in the graph
14:21:59 [danb_lap]
...so some case to standardise a non-xml syntax for such abstract queries
14:22:11 [danb_lap]
...obviously good to have an XML version of that (compare xquery, which did both)
14:22:28 [danb_lap]
...also case for some overlap w/ rule stuff eg ruleml (?missed point)
14:22:33 [danb_lap]
...abstract syntax pretty common
14:22:46 [danb_lap]
...engines differ a lot re the kind of inferences they do underneath
14:22:51 [danb_lap]
...but the querying language loioks the same
14:23:03 [danb_lap]
...you query notional, possibly infinite, dataset
14:23:18 [danb_lap]
...like simpler if we make thw QL simpler, leave service capabilities a separate problem
14:23:32 [danb_lap]
jos: you said Fwd inference
14:23:43 [danb_lap]
timbl: conceptually you are querying all the possible derrived data
14:23:50 [danb_lap]
...but we're covering that up
14:23:52 [danb_lap]
...i looked at ruleml
14:24:02 [danb_lap]
....it has a language which has an xml notation
14:24:12 [danb_lap]
...original goal to unify all tyhe various rule systems out there
14:24:20 [danb_lap]
...typically those weren't webized, ie. use URIs
14:24:33 [danb_lap]
....ruleml now extended to be URI capable
14:24:53 [danb_lap]
...can convert eg RQL and ruleml, but if rules lack uris, you have to go add namespace uris etc before get interop
14:25:00 [danb_lap]
....so ruleml was extended
14:25:08 [danb_lap]
...u/stand theres a later unpublished version
14:25:24 [danb_lap]
(aside from danbri: http://www.w3.org/2003/03/rdfqr-tests/ )
14:25:36 [danb_lap]
timbl, there are dtds to be found on ruleml site
14:25:49 [danb_lap]
...another difference: some languages allow a template for what you want returned
14:25:54 [danb_lap]
...this looks like a template language
14:26:03 [danb_lap]
(er sorry, entailment language, or something)
14:26:19 [danb_lap]
...if you think of these as languages sent to services, can see result as bindings versus an rdf graph
14:26:36 [danb_lap]
timbl shows his table of pro/cons re graph vs query, see http://www.w3.org/2001/sw/meetings/tech-200303/query
14:26:53 [danb_lap]
...mentions poss of generating an alternate(???)
14:27:00 [danb_lap]
...if you think of more pro/con, let me know
14:27:08 [danb_lap]
...i tihnk you probably need both
14:27:15 [danb_lap]
...there are also syntax choices
14:27:18 [danb_lap]
...SELECT yadda
14:27:25 [danb_lap]
...sort of thing we do in a WG
14:27:33 [danb_lap]
...arguing round brackets vs square ones
14:27:46 [danb_lap]
...semantics of what happens underneath more of less orthogonal
14:27:51 [danb_lap]
...deciding on builtins again orthogonal
14:27:56 [danb_lap]
...has happened already in various places
14:28:05 [danb_lap]
...would make sense to pick up XQuery work on this
14:28:21 [danb_lap]
...Steve Reed from Cyc has a list of builtins they support, and those of xml query, compare/contrast
14:28:26 [danb_lap]
(@@Url/google anyone)
14:28:35 [danb_lap]
...have to pick your favourite libraries
14:28:53 [danb_lap]
...will be services that do/don't support these
14:29:06 [danb_lap]
....we can make intelligent systems that add such things to dumber stores
14:29:11 [danb_lap]
...didn't get into remote query
14:29:19 [danb_lap]
...main reason we do stuff at w3c network centric
14:29:30 [danb_lap]
...one way to do this is wrap the query, uri-encoded, and append it to an http URI
14:29:36 [danb_lap]
...simple way of query, eg with a '?"
14:29:44 [danb_lap]
...easy way of attaching to existing servers
14:29:50 [danb_lap]
...or else do it one way or another in soap
14:29:58 [danb_lap]
...some arbitrary design choices there
14:30:05 [danb_lap]
...if i had to think of possible work, this is what came up
14:30:14 [danb_lap]
...i would imagine that a wg would aim for an abstract syntax
14:30:38 [danb_lap]
see http://www.w3.org/2001/sw/meetings/tech-200303/query 'possible deliverables' section
14:30:46 [danb_lap]
[[
14:30:47 [danb_lap]
So, if work were begun in this area, formally or informally, more or less in chronological order, one might hope to see:
14:30:47 [danb_lap]
Abstract syntax of query language - probably described in RDF.
14:30:47 [danb_lap]
Definition of a few conformance levels (monotonically increasing in features supported)
14:30:47 [danb_lap]
A common concrete syntax in compact (non-XML) form
14:30:48 [danb_lap]
Ontology for description of inference services provided by a service.
14:30:51 [danb_lap]
A set or sets of standard functions
14:30:52 [danb_lap]
A profile or profiles which combine the above to enhance interoperability, when experience with common engines is sufficient to define interop levels.
14:30:56 [danb_lap]
]]
14:31:02 [danb_lap]
...inference service profiles, so you 'know what you're getting'
14:31:07 [AndyS]
A concrete syntax needed for network use between machines
14:31:32 [danb_lap]
[timbl talks through bullets above]
14:32:10 [danb_lap]
...remote query: could be trivial, or ratholes
14:32:27 [danb_lap]
(i think lots of work there -- jdbc etc have lots of admin features, danbri)
14:32:45 [danb_lap]
timbl, adding DELETE etc would make a bigger job
14:32:53 [danb_lap]
...how much work, is it time to do it as a WG?
14:33:18 [alberto]
alberto has joined #swarch
14:33:30 [danb_lap]
...any general feelings re scope?
14:33:41 [danb_lap]
Harold: do you have general notion of queries that encompass rules?
14:33:49 [libby]
hey alberto!
14:33:56 [danb_lap]
timbl: if you look at the rdf query, there is a rule language that is very connected to it
14:34:03 [danb_lap]
(@@url for logs for alberto to read?)
14:34:11 [RRSAgent]
See http://www.w3.org/2003/03/07-swarch-irc#T14-34-03
14:34:11 [alberto]
hello - sorry I am late :)
14:34:13 [danb_lap]
...clearly there is a connection
14:34:20 [alberto]
ok, going through it now...
14:34:39 [danb_lap]
...for eg. RQL examples could all convert into ruleml
14:34:47 [danb_lap]
harold: rules are slightly more general concept than queries
14:34:56 [danb_lap]
...can chain queries as a rule
14:34:59 [alberto]
alberto has joined #swarch
14:35:06 [danb_lap]
...woudln't speak really of a chain of queries
14:35:20 [danb_lap]
...could call it rules and query, or really just rules
14:35:26 [danb_lap]
timbl: we have www-rdf-rules list
14:35:28 [PStickler]
I wouldn't call all chains of queries rules
14:35:32 [danb_lap]
...some folk demanded separate mailing lists
14:35:43 [danb_lap]
...i agree they're basically the same thing, but there are different systems, engines
14:36:00 [danb_lap]
harold: from last night, ruleml folk promised to submit more discussions to www-rdf-rules
14:36:09 [danb_lap]
beng: what would be way to distinguish non-query rules?
14:36:15 [danb_lap]
...a lot of rule systems run fwd
14:36:22 [danb_lap]
...query is pretty much backward
14:36:49 [danb_lap]
...if you have procedural attachment, actions, that extends beyond basic semantics of query
14:36:51 [danb_lap]
(?)
14:36:57 [danb_lap]
...two aspects that go behyond basic query
14:37:08 [danb_lap]
...from a tech pt of view, there remains a v close relationship
14:37:19 [danb_lap]
...you'd want same semantics
14:37:35 [danb_lap]
...can have simple stateless way to define rules mechanics
14:37:37 [danb_lap]
(?)
14:37:43 [danb_lap]
...v closely associated w/ a pure belief view
14:37:56 [danb_lap]
...storing queries, having queries built from subqueries
14:38:11 [danb_lap]
...wasn't emphasised
14:38:18 [danb_lap]
...should think about when is the right time to get into that
14:38:28 [danb_lap]
timbl: some people mentioned desire to store queries
14:38:31 [RalphS]
RalphS has joined #swarch
14:38:41 [danb_lap]
...one motivation for sending rules w/ a query is to add
14:38:42 [danb_lap]
(?)
14:38:47 [sanScribe]
sanScribe has joined #swarch
14:38:59 [danb_lap]
ianh: from a tech pt of view, rules in general are nothing diff from what we have in the onto languages
14:39:12 [danb_lap]
...std axioms we have in ontology, eg for subclass, is just a rule
14:39:30 [danb_lap]
...but onthe other side, query languages normally have this special feature that you only get back answers w/ fininte set of things
14:39:56 [danb_lap]
...when you say that the answer to a rule isn't all poss conclusions, it is just concrete answers
14:40:02 [danb_lap]
...eg 'tell me all the people that live in ...'
14:40:11 [las]
las has joined #swarch
14:40:19 [danb_lap]
...that gives a completely diff computational properyty to the language
14:40:23 [danb_lap]
timbl: does it change the syntax?
14:40:36 [danb_lap]
...can we have different operationals but keep the ql the same
14:40:41 [danb_lap]
ian: maybe...
14:40:47 [sandro]
sandro has joined #swarch
14:40:53 [danb_lap]
...my point was that we might allow some things in a QL that we don't allow in an assertional language
14:41:06 [danb_lap]
...since QL resultsets have different characteristics
14:41:27 [danb_lap]
timbl: [draw attention to ian, ben ... work on OWL<->Rule mapping] @@url?
14:41:40 [danb_lap]
frank: in QL, constructing new tuples
14:41:53 [danb_lap]
...would want a ql to support certain kinds of construction
14:42:06 [danb_lap]
ian: true, but elements in the tuples are things we know about in advance
14:42:28 [danb_lap]
[...]
14:42:39 [danb_lap]
frank: we need to be clear about the extent to which this is a restrictuion
14:42:51 [danb_lap]
...if we have cities example, links between cities
14:42:58 [danb_lap]
...q is: am i constructing paths
14:43:04 [danb_lap]
...perfectly reasonable thing to do
14:43:18 [danb_lap]
ian: infinite answer in general, as can to/fro the same pair of cities
14:43:34 [danb_lap]
...generally you disallow such queries, by saying 'gimme acyclic paths'
14:43:50 [danb_lap]
frank: is notion of a path, or instances of paths, what you're considering old things vs new things
14:43:50 [danb_lap]
(?)
14:43:56 [shellac]
shellac has joined #swarch
14:44:00 [danb_lap]
timbl: you may be talking about 'path' differently
14:44:31 [danb_lap]
http://www.w3.org/2003/03/07-swarch-irc
14:44:46 [danb_lap]
ian: if you allow infinte paths and infinite poss answers, thats when the wheels fall off
14:44:58 [danb_lap]
frank: am just trying to clarify what kinds of things we can get back
14:45:01 [danb_lap]
(he isn't)
14:45:04 [libby]
sandro - no
14:45:06 [danb_lap]
(shellac i mean)
14:45:23 [danb_lap]
patrick: [...] if you havd a Q engine without rules, inference, you get back just ground stuff
14:45:33 [danb_lap]
...if it does have such support, you get things back that are implicit
14:45:42 [danb_lap]
(photos - good idea)
14:46:10 [danb_lap]
patrick: whether underlying engine has inference is separable
14:46:21 [danb_lap]
...what you get back isn't necc asserted/explicit in your kb
14:46:32 [danb_lap]
ian: i wasn't intending to say what ought be in/out of language
14:46:48 [danb_lap]
...just note that computational properties of a ql vs an assertional language differ
14:46:57 [danb_lap]
...because of this fact that you know in advance finite set of answers
14:47:18 [danb_lap]
timbl: proposal is that you can use the same QL in both contexts (albeit w/ diff comp properties)
14:47:26 [danb_lap]
...use same lang to talk to it, results come back the same
14:47:45 [danb_lap]
...hypothesis is that the ql can be the same
14:48:00 [libby]
DanC: this sort of thing? http://swordfish.rdfweb.org/discovery/2002/10/calwk/calwk.jsp?pics=+calendar+workshop+images+ (slow...)
14:48:08 [DanC]
Volz: ...
14:48:46 [danb_lap]
timbl: qls i saw didn't allow (various kinds of) fancy path ql
14:49:04 [DanC]
spiffy, libby. that's more than I had in mind, but that's cool.
14:49:32 [danb_lap]
pathayes: here when you say Query Lnaguage, are you talking about the abstract pattern, or all the other additional features too
14:49:32 [PStickler]
If a given query engine is not able to answer 'true' that doesn't mean the answer is 'false'
14:49:39 [bwm]
volz said he things query language should support expression of rules that can be used for inference - at least thats what I heard
14:49:59 [PStickler]
If one engine does not have inference, and the answer is implicit, then it simply cannot say, but another engine with inference may be able to answer positively
14:50:05 [danb_lap]
timbl: ... [summaries earlier discussion for pat's benefit]
14:50:16 [danb_lap]
bwm, maybe later so i can say stuff! ok for now.
14:50:23 [DanC]
hmmm... the calwk photos aren't all from the workshop. I'm thinking of photos that give evidence that, e.g., I was here.
14:50:56 [alberto]
yes very useful work danb
14:51:28 [danb_lap]
ericp: [...] fact that you have a rule engine that might stop on 1st answer... is characteristic of the engine ... not of the engine
14:51:38 [danb_lap]
(some discussion about whether eric had this backwards)
14:51:41 [danb_lap]
s/engine/rule/
14:52:10 [danb_lap]
...if we can keep in mind there's a large commonality, and that you can express things as characteristics of the query or the report
14:52:22 [danb_lap]
(timbl is doing things on paper)
14:52:36 [DanC]
I don't expect "I want the first answer only" to be part of the query language syntax. hmm... .
14:52:42 [danb_lap]
jos: w.r.t. resultset, i think the results as a primitive form of proof/explanation in rdf, is wortyh considering in our experience
14:52:46 [danb_lap]
me neither, danc
14:53:00 [danb_lap]
jos: ...worth explaining in an rdf graph, facts/rules/queries/proofs all part of this
14:53:13 [danb_lap]
...proofs as continuations, could be given to another engine to continue processing
14:53:23 [danb_lap]
timbl: this is a list of other poss extensions
14:53:40 [danb_lap]
(@@we should transcribe later. Have 'saving result set'; returning proof; so far)
14:54:18 [danb_lap]
brian: way we're looking at this in jena
14:54:22 [danb_lap]
..rel w/ rules v query
14:54:35 [danb_lap]
..you target a query at an engine, you identify the rules you want to apply by picking a target
14:54:45 [danb_lap]
...implicit suggestion that the two differ
14:54:58 [danb_lap]
...but QL could have a hole saying 'and some companion rules could go here'
14:55:57 [danb_lap]
danbri (barging in): this suggets a way that work can be punted from QL to companion APIs
14:56:25 [danb_lap]
ian: just to try to clarify point about computation, same thing raphael was saying...
14:56:29 [DanC]
:flipA log:semantics { :OtherFeatures is rdf:type of :savingResultSet, :returnedProof }.
14:56:43 [danb_lap]
...is perfectly possible to take OWL and guarantee you can compute answers to all quewries in that lang
14:56:47 [danb_lap]
...ie complete
14:57:03 [danb_lap]
...same language as an assertional language, can show it isn't complete
14:57:09 [danb_lap]
(@@did i get that right?)
14:57:25 [DanC]
:flipA dc:description "(timbl is doing things on paper)".
14:57:31 [danb_lap]
...you may not care about completeness, but point is you get diff comp properties
14:57:38 [PStickler]
Query and Rule need not be different languages, but rather, Query is a subcomponent of the Rule language
14:57:47 [danb_lap]
harold: ... you can perhaps have two kinds of queries, lookup vs inferential
14:58:00 [danb_lap]
...subqueries, could be lookup or further composite
14:58:12 [danb_lap]
...in ruleml, we have an xml element 'query' with flags for lookup vs inference
14:58:27 [danb_lap]
timbl: thinks thats a shared model
14:58:38 [danb_lap]
mikedean: one big question... how much do q and rules overlap
14:58:42 [danb_lap]
...am trying to do a venn diagram
14:58:54 [danb_lap]
...what i have not quite ready to share but suggest useful technique
14:59:02 [danb_lap]
...prelim, that most stuff is out on the edges
14:59:12 [danb_lap]
...suggest interesection is pretty small
14:59:23 [danb_lap]
(@@is that good? ie. a small focussed workitem?! --danbri)
14:59:41 [danb_lap]
path: there is a std logical picture of the query process
14:59:46 [danb_lap]
...and how query/rules overlap
14:59:54 [danb_lap]
...rules are like horn clause implication
15:00:05 [danb_lap]
...a query is a pattern put up as a candidate conclusion
15:00:07 [danb_lap]
(?)
15:00:10 [danb_lap]
...
15:00:10 [PStickler]
Specifying whether a given engine does or does not perform inference is a parameter to the engine, not a feature of the core query language
15:00:20 [danb_lap]
...get this on table as the 'off the shelf' picture of the rel'nship
15:00:27 [danb_lap]
...also Jos's point about returning proofs
15:00:50 [danb_lap]
...nice story about poss responses, range from 'yes'! to getting entire proof, versus intermemdiate, getting bindings
15:00:55 [PStickler]
I am in the queue after DanBri (though got skipped over already once)
15:00:57 [danb_lap]
...might well be other things
15:01:01 [danb_lap]
q-
15:01:18 [danb_lap]
zakim, DanBri is danb_lap
15:01:20 [Zakim]
sorry, danb_lap, I do not recognize a party named 'DanBri'
15:01:39 [DanC]
ack Pat
15:01:39 [danb_lap]
ben: to follow that up, jos and pat touched on proof
15:01:45 [DanC]
ack Ben
15:01:47 [danb_lap]
...query as a concept in KR is something that any KR can have
15:02:04 [PStickler]
I am in the queue on the whiteboard
15:02:05 [danb_lap]
...you usually start conceptually from a KR, eg principles of sanctioned entailment
15:02:25 [danb_lap]
w/ proof, other actions, consistency, monotoniciity, syntactic violations, resource limits, max answers etc.
15:02:34 [danb_lap]
...mechanical or complemenetary surrounding considerations
15:02:44 [danb_lap]
...much of xquery focussed on such things
15:02:53 [danb_lap]
timbl: did xquery cover resource limits?
15:03:03 [danb_lap]
ben: eg don't try more than 1000 seconds on this
15:03:19 [danb_lap]
...if you doing info integration across sites, spend only so much money/time
15:03:25 [danb_lap]
timbl: didn't think this was in xquery
15:03:44 [DanC]
ack PatrickS
15:03:44 [danb_lap]
ben: not all in xquery, but similar concerns
15:04:00 [danb_lap]
patricks: (i) there needn't be two different languages, query vs rule
15:04:05 [danb_lap]
...q can be subcomponent of Rule
15:04:19 [danb_lap]
...in add to what you just got back from this q, here do some things
15:04:32 [danb_lap]
(ii) should distinguish the QL versus rest of msg you're communicating to server
15:04:49 [danb_lap]
(lists some practical stuff as ben did above, eg. how much server resource to spend)
15:04:58 [danb_lap]
...additional component, what you do with the results
15:05:08 [danb_lap]
timbl: these things don't have to clutter up the QL
15:05:09 [DanC]
ack bwm
15:05:29 [danb_lap]
brian: i wish andys was here! re harolds point about query decomposition... should note that Qs often go remotely
15:05:41 [danb_lap]
...so processing model isn't necc that of a low-latency API
15:05:49 [danb_lap]
...often want to get bigger chunks due to net
15:06:20 [danb_lap]
...process issue: when it comes to doing some WG-ish work, we need to start w/ basics first, get something simple running first
15:06:22 [danb_lap]
(danbri claps_)
15:06:32 [PStickler]
Having chains of queries doesn't mean that each subquery in the chain is executed between client and server independently, rather the entire chain can be specified and passed to the engine to process
15:06:34 [danb_lap]
timbl: so if we restricted rule-oriented work in first phase...
15:06:35 [libby]
+1 bwm
15:06:46 [danb_lap]
...eg don't do anything at this stage that we don't need for query
15:06:59 [danb_lap]
brian: i wouldn't go that far
15:07:07 [danb_lap]
...just when chartering, emphasise simple/quick/soon
15:07:23 [danb_lap]
danc: counterpoint to mike's point that intersection is smaller than union
15:07:31 [danb_lap]
...does that advance the state of the art
15:07:34 [danb_lap]
...i think it does
15:07:41 [danb_lap]
[break schduled in 25 mins]
15:07:50 [danb_lap]
zakim, remind me in 25 minutes to break
15:07:51 [Zakim]
ok, danb_lap
15:08:05 [DanC]
ack timbl
15:08:15 [DanC]
ack DanC
15:08:18 [danb_lap]
timbl: [talks about bindings... vs graph]
15:08:45 [danb_lap]
timbl: likely on server, many things happening, going over web, inference etc
15:09:00 [danb_lap]
...so affects return proof, complexity
15:09:06 [danb_lap]
...but i don't see that changing the query language
15:09:23 [danb_lap]
em: swordfish folks, pls follow up on brian's pt
15:09:33 [danb_lap]
randy: i feel brian's point well taken
15:09:42 [danb_lap]
(Randy from Sun)
15:09:45 [danb_lap]
em: you're doing rdf query
15:09:52 [danb_lap]
randy: we're pulling triples really, not full query
15:10:01 [danb_lap]
[speaker?] [missed point]
15:10:11 [danb_lap]
path: how are you handling bnodes?
15:10:16 [danb_lap]
timbl: do you have a ql?
15:10:25 [danb_lap]
[?]: no, we tried versa, temporarily
15:10:47 [libby]
? is sudeep something
15:10:51 [danb_lap]
timbl: cc/pp WG going into CR phase, lang for describing device capacities
15:11:05 [Nobu]
Nobu has joined #swarch
15:11:10 [danb_lap]
...data is there, and thy're generating SVG-with-SMIL or whatever, as a function of a piece of RDF
15:11:36 [danb_lap]
timbl: they want a .js api
15:11:40 [danb_lap]
...i'm told mozilla does this
15:11:42 [DanC]
ack Harold
15:11:45 [danb_lap]
danbri: it does, they have a full rdf api
15:11:49 [DanC]
ack EricM
15:12:00 [danb_lap]
harold: views in rdbms have always been rules, datalog like
15:12:15 [danb_lap]
...since sql99 allows recursive rules
15:12:23 [danb_lap]
...i think we shouldnt' go behind sql99
15:12:41 [danb_lap]
...in yr table tim, already classical db langauges have datalog rules
15:12:52 [danb_lap]
timbl: you mean we shouldn't lag behind sql folks
15:13:00 [DanC]
a webized syntax for datalog is what I think is ripe.
15:13:07 [danb_lap]
[missing cople points from tim]
15:13:16 [danb_lap]
harold: new work would be reducing a query to subqueries
15:13:40 [danb_lap]
...looking at rules from bottom up(?)
15:13:57 [danb_lap]
frank: Is concurrency at all within scope of this activity
15:14:10 [danb_lap]
timbl: i'd say out of scope for rule language, in scope for SOAP
15:14:19 [danb_lap]
...web services will allow ways of composing web services
15:14:26 [danb_lap]
(locks, atomic ops etc i guess --danbri)
15:14:53 [danb_lap]
...i put up 'profiling', eg we coudl say 'everyone using R*QL in practice will also need to agree following behaviours'
15:15:17 [danb_lap]
frank: an addon... in a general siutaiont i might want to send an ontology based meta description re concurrency
15:15:26 [danb_lap]
timbl: you mean, meta info about the query...?
15:15:36 [danb_lap]
frank: std assumption is serializability
15:15:55 [danb_lap]
...you can imagine describing in a concurrency ontology, richer details
15:16:27 [danb_lap]
ericp: i think the (my?) QL survey paper has some stuff that grounds this
15:16:38 [danb_lap]
timbl: i should mention that my summary based on three summary papers
15:16:59 [bwm]
tim brings up eric p's page
15:17:14 [danb_lap]
see http://www.w3.org/2001/11/13-RDF-Query-Rules/
15:17:15 [bwm]
url to follow
15:17:27 [bwm]
what looks like an xml structure
15:17:30 [bwm]
a language binding
15:17:40 [bwm]
match has characteristics
15:17:46 [bwm]
as do report and bindings
15:18:27 [bwm]
query languages characterised by characteristics of various components of the language
15:18:46 [bwm]
identifying components and characteristic ontology is approach to design
15:20:02 [bwm]
bwm asks if update is in scope
15:20:09 [bwm]
timbl says no
15:20:42 [bwm]
timbl would like keep update off critical path - we don't have 10 update languages yet
15:21:46 [bwm]
pats: comment to eric's slide; 3rd option - ask for bindings, ask for subgraph, ask for graph of all you found
15:22:12 [bwm]
pats: doesn't see that chaining of queries is necessarily a rule
15:22:55 [bwm]
pats: do one query, then rank those results, then project
15:23:05 [bwm]
timbl: no one has mentioned ranking
15:23:12 [bwm]
timbl: does the query language need ordering
15:23:17 [bwm]
timbl: problem for rdf
15:23:33 [bwm]
pats: that is not part of a query language, its metadata about the query
15:24:25 [bwm]
ben: ordering is very useful in general especially for broad area web query
15:24:40 [bwm]
ben: you have to be careful, not to be alittle bit pregnant
15:24:47 [danb_lap]
(hmm ordering seems pretty close to datatyping issues...)
15:24:53 [bwm]
ben: you can't do mathcing "more or less closely" on the cheap
15:25:17 [bwm]
specifying the ntion of closeness is really very different from ranking
15:25:41 [bwm]
if we incorporate it we pull in techniques from info retrieval, baysian reasoning, fuzzy reasoning
15:25:42 [PStickler]
You can express abstractions of ordering without having to specify how each engine actually calculates ordering
15:26:05 [bwm]
pats: its not opart of the query langauge, its metadata about the query
15:26:13 [bwm]
pats: different engines may order things differently
15:26:25 [bwm]
pats: may choose engine that is appropriate
15:26:57 [bwm]
lynn: ordering in a query result set is something that we will need and its important, but its different from the fundamental query langauge
15:27:15 [bwm]
lynn: the reason you need ordering is that the things you order higher are better
15:27:24 [bwm]
lynn: doesn't belong in this query language
15:27:24 [PStickler]
Goodness can simply be percentage of partial match
15:27:52 [bwm]
lynn: its a differnt thing to ask a quyery that has a yes/no answer
15:28:05 [danb_lap]
(i think each match can be 100% good, yet we still operationally want some of them first, eg for UI generation reasons -- mozilla have some use cases here, see XUL)
15:28:26 [bwm]
lynn: if we do goodness of fit it will make simple matching must harder
15:29:05 [bwm]
lynn: build langauge with binary answers but keep in mind that we will build an infrastructure on top that will do richer things
15:29:16 [bwm]
sandro: I thought we were talking about ordering not goodness of fit
15:29:37 [bwm]
sandro: wants ordering, but no lynn's type of ordering
15:29:41 [bwm]
s/no/not/
15:29:41 [PStickler]
I used the term 'ranking' which has to do with goodness of fit, or completeness of match
15:29:51 [las]
It's important to have (and to understand that sweb will have) imprecise matching. It's just not the same as the first query language.
15:30:27 [bwm]
harold: wants to support sandro: ordering is a kind of aggregate which could be built into a rules language
15:30:32 [PStickler]
Ordering/ranking is not part of the query language, but is part of the query solution, and is communication to the engine just as requests for bindings or proofs as results
15:30:47 [bwm]
harold: its like applying a built in afger a query
15:31:23 [bwm]
ericP': report characteristic - gives back all triples that have proved themselves - triple closure
15:31:34 [bwm]
... can nest them like nested sql statements
15:31:39 [bwm]
... look for best case
15:31:48 [bwm]
... give me back the first answer, not exhaustive
15:32:04 [bwm]
... that would handle scenario without aggregates or ordering
15:32:22 [bwm]
frankm: when the quetsion of ordering came up, can we produce an ordered thing (whatever the ordering)
15:32:32 [bwm]
... i.e. can we produce an rdf:Seq
15:32:40 [bwm]
... what kind of closuer does the query language have
15:32:51 [bwm]
... can it produce all the collection types in RDF
15:32:51 [Zakim]
danb_lap, you asked to be reminded at this time to break
15:33:06 [bwm]
... there is no word for that set of triples
15:33:36 [bwm]
path: in the matter of order its vital that if queries can specifiy order, the engine should be able to ignore it
15:33:41 [DanC]
rev 1.40 of mtg page has lightning talks. http://www.w3.org/2001/sw/meetings/tech-200303/
15:33:50 [bwm]
... invite randy to say more about the other issue
15:33:54 [bwm]
... which is update
15:34:28 [bwm]
pats: agrees fully with pat
15:35:17 [bwm]
ben: its important to distinguish between semantically generated notion of exact, better best, and the ability to manupulate ordered lists as concepts in the langauge
15:35:26 [bwm]
... most rule languages have that sort of thing
15:35:45 [bwm]
... that sort of ordereredness is something we all agree we want to have
15:36:03 [bwm]
... disagrees with distrinction that patS is trying to make
15:36:15 [bwm]
miked: people will want sql like ORDERED clause
15:36:27 [PStickler]
I am speaking about RANKING not ORDERING
15:36:31 [bwm]
... thinks that is something that is not really in a rules language
15:36:40 [bwm]
whose speaking
15:37:02 [libby]
rand possibly?
15:37:03 [ericP]
Rand is speaking
15:37:07 [bwm]
??? says query language gives you an ordering mechansim
15:37:14 [bwm]
... other ordering is higher order
15:37:24 [bwm]
timbl: we have collections
15:37:37 [bwm]
... there are time when we want to do aggregations
15:37:58 [bwm]
... before you can say something is closest requires a closed world
15:38:04 [PStickler]
We need to differentiate between ORDERING, which is an operation based on the results of a query, and RANKING which is an operation performed during the execution of a query
15:38:19 [bwm]
...ordering requires a closed world
15:38:38 [bwm]
... need formulae
15:38:49 [bwm]
... otherwise the rule makes no sense
15:39:07 [bwm]
scribe isn't following this
15:39:19 [bwm]
timbl: pats can have last word
15:39:32 [bwm]
pats: differentiate between ordering and ranking
15:39:43 [bwm]
... ordering is about sorting the results of a query
15:39:50 [bwm]
... ranking affects the query itself
15:40:16 [bwm]
... metadata about the query is separate from the query itself
15:40:32 [bwm]
ajournded for a short time
15:40:44 [bwm]
lightening talks - the timer is non negotiable
15:41:00 [bwm]
break for 5 mins
15:47:29 [DanC]
q= Stickler, Volz, Miller, Grosof, Boley, Tabet, Brickley, Sudeem, McBride, De Roo, Hawke
15:47:40 [DanC]
Zakim, give each speaker 5 minutes
15:47:41 [Zakim]
I don't understand 'give each speaker 5 minutes', DanC
15:48:07 [DanC]
Zakim, allow each speaker 5 minutes
15:48:08 [Zakim]
ok, DanC
15:49:20 [Nobu0]
Nobu0 has joined #swarch
15:49:26 [bwm]
zakim, has more power than I realised
15:49:27 [Zakim]
I don't understand 'has more power than I realised', bwm
15:51:47 [DanC]
q= Stickler, Volz, Miller, Boley, Grosof, Tabet, Brickley, Sudeem, McBride, De Roo, Hawke
15:53:50 [DanC]
ack Stickler
15:53:55 [timbl]
timbl has joined #swarch
15:56:44 [libby]
erk
15:57:28 [Zakim]
Zakim has joined #swarch
15:58:59 [timbl]
I have pointed out on #RDFIG before that MGET is counter to Semantic Web architectrue and Web architecture.
15:59:21 [DanC]
ack Volz
16:00:10 [RalphS]
KAON
16:00:59 [RalphS]
Raphael Volz presents KAON
16:01:39 [timbl]
ref: Graphlog, Mendelson (sp?) Univ toronto [PODS90]
16:02:58 [JosD___]
JosD___ has joined #swarch
16:03:01 [Zakim]
Zakim has joined #swarch
16:05:01 [Zakim]
Zakim has joined #swarch
16:05:13 [RalphS]
queue=miller
16:05:35 [DanC]
started
16:05:40 [PStickler]
PStickler has joined #swarch
16:06:11 [alberto]
alberto has left #swarch
16:06:50 [RalphS]
Libby Miller speaks about Query testcases
16:08:30 [RalphS]
DanBri: please mail testcases to www-rdf-rules
16:10:43 [DanC]
Boley started
16:10:53 [timbl]
ack Boley
16:11:04 [libby]
all the info for that 'talk' is in http://www.w3.org/2003/03/rdfqr-tests/, http://lists.w3.org/Archives/Public/www-rdf-rules/2003Feb/0014.html
16:16:12 [simonMIT]
simonMIT has joined #swarch
16:16:18 [chaalsBOS]
chaalsBOS has joined #swarch
16:16:37 [DanC]
Grosof started
16:17:18 [DanC]
ACTION Boley: contribute presentation materials to meeting record
16:17:27 [DanC]
ACTION Volz: contribute presentation materials to meeting record
16:18:51 [DanC]
I see 20030217-outline.html in the addressbar of what's projected
16:19:27 [alberto]
alberto has joined #swarch
16:20:19 [libby]
yeah it starts with C: or somethign though - I was trying to get it down last night.
16:22:01 [DanC]
Tabet started
16:22:34 [timbl]
Ben said that the document he just showed in on teh Join Committee archive
16:24:27 [DanC]
ACTION Tabet: contribute presentation materials to meeting record
16:24:35 [timbl]
I am surprised to see CWM described as a RuleML application, And wonders how many if any of the others on the list actually use RuleML!
16:27:55 [DanC]
danbri started
16:28:14 [libby]
http://www.foafnaut.org
16:33:09 [libby]
a javascript and SVG demo which queries RDF data about people
16:33:48 [DanC]
Sudeed (sp?) started
16:39:07 [jhendler_]
jhendler_ has joined #swarch
16:39:25 [DanC]
q?
16:39:35 [RalphS]
Bernard: would like a query language to be useable without requiring the user to understand the entire model
16:40:25 [DanC]
ack McBride
16:41:07 [DanC]
McBride: not talking about 'using RDF to track group work' as advertised
16:41:23 [em-lap]
rdf button...
16:41:48 [em-lap]
http://www.w3.org/RDF/icons/rdf_flyer
16:42:06 [em-lap]
http://www.w3.org/RDF/icons/rdf_flyer.24 for the size BrianM is looking for
16:42:53 [DanC]
hmm... restaurant service... gotta tell dwj...
16:42:55 [DanC]
djw
16:43:58 [timbl]
Hmmmm.. I must talk with Brian about the file system ontology as I want to add cwm builtins.
16:44:04 [danb_lap]
more src files from foafnaut demo: http://foafnaut.org/2003/snap1/cache-sha1/ (these are .gz'd person descriptuions)
16:45:58 [danb_lap]
brian, timbl: see http://lxr.mozilla.org/mozilla/source/rdf/datasource/src/nsFileSystemDataSource.cpp for Mozilla rdf filesystem datasource
16:46:04 [DanC]
ack De Roo
16:46:50 [GuusS]
DanC, any chance of a short demo from me on semantic image annotation? A bit lte...
16:49:13 [dfg_olin]
dfg_olin has joined #swarch
16:51:52 [DanC]
ack Sandro
16:52:29 [JosD___]
JosD___ has joined #swarch
16:52:35 [RalphS]
-> http://www.w3.org/2003/Talks/0307-semrun/ Sandro's talk
16:57:19 [danb_lap]
semrun: http://www.w3.org/2001/12/semrun/
16:57:40 [DanC]
ack Carroll
16:58:11 [RalphS]
Jeremy Carroll: Signing an Ontology
17:07:02 [libby]
I can try and get some form now
17:07:08 [libby]
josd likes testcase approach
17:07:11 [danb_lap]
brian: we (jena group) would like to see a WG, a synthesis of existing query work
17:07:25 [danb_lap]
jjc: HP has decided we would have someone on a WG
17:07:34 [danb_lap]
em: would you be willing to put rdql up as a note
17:07:38 [libby]
danbri/libby not keen on wg, like testcases etc
17:07:43 [danb_lap]
brian: dont seee why not. andy's person to ask.
17:07:56 [danb_lap]
...
17:08:10 [danb_lap]
brian: plethora of similar-but-different rdfq systems... don't see same w/ rdf rules right now
17:08:22 [danb_lap]
am nervous of taking on too much, a lesson from rdfcore
17:08:35 [danb_lap]
ben: what is the right way to structure a discussion about whether to have a wg
17:09:12 [danb_lap]
danbri: get out there on the mailing lists and show progress!
17:09:19 [danb_lap]
timbl: at some pt we need a charter...
17:10:43 [danb_lap]
timbl: if we have a handle on the architecture, we can start drafting a charter
17:10:53 [danb_lap]
harold: when the webont group was started, was an explicit decision not to treat rules
17:11:03 [danb_lap]
...was often discussed that rules might be the next wg
17:11:11 [danb_lap]
...am nervous that the next wg might only be re query
17:11:19 [danb_lap]
...that for 2nd time, rules postponed
17:11:34 [danb_lap]
said: add one comment. If you want RDF to industry, it is important to start the work.
17:11:43 [danb_lap]
...instead of saying 'lets do queries first'
17:11:52 [danb_lap]
em: 2nd danbri's point
17:12:02 [danb_lap]
...yes when charter was written we tried to bound OWL work
17:12:06 [danb_lap]
...criticial to scope work
17:12:15 [danb_lap]
...charter also said 'this will be done in 8 months'
17:12:27 [danb_lap]
...things tend to take longer when hit edge cases
17:12:38 [danb_lap]
...case has to be made, in terms of industry adoption, individuals,
17:12:50 [danb_lap]
...case is still being made
17:13:01 [danb_lap]
danc: a strategy used succesfully w/ xml
17:13:06 [danb_lap]
...put a long roadmap in place
17:13:12 [danb_lap]
...jon bosak wrote xml activity statement
17:13:26 [danb_lap]
...wg produced xml 1.0 after 18 months, then activity split into pieces to finish the work
17:13:39 [danb_lap]
...strategy is used a lot, to say 'come join, then we'll do bits at a time'
17:13:53 [danb_lap]
frank: seconding harold's nervousness about considering queries without rules
17:14:08 [danb_lap]
...for the SW apps of ontology langauges, i'm lost without ability to specifiy rules
17:14:09 [alberto]
alberto has joined #swarch
17:14:30 [danb_lap]
path: i'd like disagree... one can rationally consider queries without rules
17:14:38 [danb_lap]
...and then put them together
17:14:51 [danb_lap]
em: frank, yes v important. There is a list, www-rdf-rules, that people look to for that discussion
17:15:33 [danb_lap]
...if you have these concerns, ...
17:15:44 [danb_lap]
danbri: part of making the case for a WG, is finding internal-to-w3c customers, other WGs
17:16:00 [danb_lap]
patricks: re rules, would hope rule folk would participate to ensure a rule-ready ql
17:17:10 [danb_lap]
adjourned for lunch.
17:17:42 [danb_lap]
--------------
17:25:13 [aaronofmo]
aaronofmo has joined #swarch
18:07:21 [libby]
libby has joined #swarch
18:25:02 [libby]
libby has joined #swarch
18:28:44 [shellac]
shellac has left #swarch
18:31:13 [chaalsBOS]
chaalsBOS has joined #swarch
18:33:09 [DanC]
coming back after lunch
18:33:13 [PStickler]
PStickler has joined #swarch
18:34:14 [DanC]
DanC has changed the topic to: SemWeb Arch in Boston http://www.w3.org/2001/sw/meetings/tech-200303/
18:36:08 [bwm]
bwm has joined #swarch
18:36:24 [DanC]
EricM convenes...
18:36:42 [DanC]
[[ 13:30 - 15:00 First afternoon session
18:36:42 [DanC]
Best Practices / Education and Outreach - Eric Miller ]]
18:37:51 [danbri]
danbri has joined #swarch
18:38:06 [DanC]
EricM introduces Kathy McDougal of Sun.
18:38:54 [danbri]
kathy: will show you one application of rdf... not claiming sun espouse rdf across the board
18:39:20 [danbri]
...how did we arrive at rdf? Knowledge Technologies. Met Uche from Forethought. Located near our Sun offices.
18:39:28 [danbri]
...walked away w/ a focus on both RDF and on DC
18:39:35 [danbri]
...a year of education w/ Uche
18:39:42 [danbri]
...talking to EricM too
18:39:51 [DanC]
ACTION KathyM: contribute presentation materials to the meeting record
18:40:02 [danbri]
...talking here both about what we're done, and about future work ideas, best practice etc
18:40:25 [danbri]
...our group, Knowledge Services, within Sun. We try to create and share knowledge to solve service issues.
18:40:29 [em-lap]
em-lap has joined #swarch
18:40:40 [danbri]
intros: melissa, randy, sandeep
18:41:08 [danbri]
...goals: help engineers on phone w/ customer support, help online self service, and ultimately avoid having problems in the 1st place!
18:41:14 [timbl]
timbl has joined #swarch
18:41:20 [danbri]
...so use rules and problem diagnosis for problem avoidance
18:41:57 [danbri]
...IC management: process by which business applies [missed]
18:42:11 [danbri]
(see slides for detail; I can't scribe all this and listen properly...)
18:42:33 [danbri]
(explanaiton that swoRDFfish -- only word with 'rdf' in it in sequence)
18:42:46 [danbri]
(no connection to swordfish.rdfweb.org btw)
18:43:19 [danbri]
slide re SunSolve site screenshot. -- more precise search.
18:43:24 [danbri]
...also content aggregation
18:43:33 [danbri]
...often traditiaonlly you need to know where knowledge lives
18:43:55 [danbri]
..eg whitepapers about products, in 5 different KBs around company
18:44:10 [danbri]
...another thing we're solving with Ontology: std names for products
18:44:47 [danbri]
......consistency across site
18:44:54 [danbri]
'progress to date' slide
18:45:47 [danbri]
- corp wide stds defined; -cross-org team formed; -exec sponsorship gained; - business education ongoing; -ontology designed and implemented; -tech infrastructure deployed; -reference implementations created.
18:46:55 [danbri]
...one of the most powerful things was to create a demo
18:47:08 [danbri]
...to show folk the value of adding metadata
18:47:54 [danbri]
...stakholders: (source systems); global sales; marketing; corporate; sun services; software.
18:48:01 [danbri]
...parts of the org we've gone out to educate
18:48:22 [danbri]
...product knowledge that we're trying to aggregate is spread around the company
18:48:29 [danbri]
sandro: how many people work at Sun
18:48:50 [danbri]
36,000
18:49:14 [danbri]
(danbri stole 'swordfish' from ericp, gave it to libby)
18:49:21 [danbri]
architecture summary:
18:49:46 [danbri]
- open standards based; sun ONE Web Server; Java RDF API; Oracle Database; JAX-RPC Web Services; N-Tier Capable
18:50:02 [DanC]
partners... yeah... fractal community...
18:50:08 [danbri]
...Standards: RDF, SOAP/XML, DAML+OIL, Java
18:50:20 [danbri]
ben: you commented earlier that you 'liked rules'; can you expand?
18:50:33 [danbri]
kathy: we do, would like to work with them more. We do configuration management...
18:50:48 [danbri]
...you can imagine someone in sun store, figiuring out what works with what
18:50:58 [danbri]
melissa: [missed example]
18:51:21 [danbri]
kathy: many other apps of rule across the business
18:51:29 [danbri]
pat hayes: same question re DAML+OIL
18:51:39 [danbri]
sandeep(?): transitive properties, some constraints
18:51:44 [danbri]
...restrictions
18:51:47 [danbri]
...not full thing
18:52:00 [danbri]
harold: do you use this at runtime, or do you pre-deduce facts
18:52:05 [danbri]
...do you do any realtime inference?
18:52:16 [danbri]
k: not realtime inference
18:52:21 [danbri]
randy: not doing much inference yet
18:52:36 [danbri]
k: we got the concept 2 years ago, spent 1st year coming up w/ the standards
18:52:52 [danbri]
getting support, data etc
18:53:03 [danbri]
ian: how do you work with the daml+oil constraints, process?
18:53:18 [danbri]
sandeep: nothing on the market, so we have something of our own in Java
18:53:27 [danbri]
(mention of a/the daml validator)(?)
18:53:35 [danbri]
mike: where do you get the data? go out to the web?
18:53:43 [danbri]
kathy: building those interfaces, getting things more automated
18:53:52 [danbri]
r: has been more hand-coded previously
18:54:02 [danbri]
melissa: as we migrate from central KB things get marked up better
18:54:15 [danbri]
...a tool consolidation effort
18:54:20 [danbri]
ian: how much data do you have?
18:54:29 [danbri]
...is it proprietary? could be useful src of tests
18:54:41 [danbri]
sandeep: we have 30k triples, mostly DC and sun product markup
18:54:49 [danbri]
...some proprietary, some opensource
18:55:11 [danbri]
randy: we have gone back and fwd with laywers and determined that we are not going to patent our schemas
18:55:19 [danbri]
(scribe note; did i get that right?)
18:55:30 [danbri]
...but that some knowledge in the schemas/ontos is pre-release
18:55:45 [danbri]
...
18:55:55 [danbri]
randy: re entitlement...
18:56:00 [danbri]
...we are a product company
18:56:16 [danbri]
...at certain pts in prod cycle, you can get to see things based on who you are
18:56:33 [danbri]
...eg 45 days pre release, peeople in certain category see certain stuff
18:56:48 [danbri]
...we are trying to use corporate ldap, not reinvent infrastructure in rdf unless needed
18:56:55 [danbri]
jos: can you comment more on web services?
18:57:16 [danbri]
sandeep: ...
18:57:20 [danbri]
...use wsdl file?
18:57:34 [danbri]
(missed detail of pt)
18:57:45 [danbri]
kathy: (returning to slides)
18:57:51 [danbri]
....recommended tags and vocab
18:57:56 [danbri]
...started with dublin core
18:58:03 [danbri]
...sub product vocab, and tech areas
18:58:27 [danbri]
(scribe note: perhaps s/sandeep/sudeep/ -- correction welcomed)
18:58:38 [danbri]
....how do we know what people we want to exchange info with are using?
18:59:00 [danbri]
....controlled vocabs: language, RFC 3066; Format, based on Mime std. Publisher. Rights. Sun Product, ...
18:59:13 [danbri]
...future opportunities
18:59:22 [danbri]
...common practices for business education
18:59:27 [danbri]
...case studies proving business impact
18:59:31 [danbri]
....registry of standards
18:59:37 [danbri]
....shared approaches to common business models
18:59:42 [DanC]
"Future Opportunities" slide
18:59:44 [danbri]
...governance model best practices
18:59:52 [danbri]
...ontology modeling best practices
18:59:56 [danbri]
...expanding tech support
19:00:15 [danbri]
- we can go out to other companies, talk about where they're seeing impact
19:00:17 [danbri]
(?)
19:00:19 [GuusS]
GuusS has joined #swarch
19:00:26 [danbri]
(re case studies)
19:00:37 [danbri]
...registry of stds being mainly vocabs,
19:00:39 [danbri]
randy: also models
19:00:46 [danbri]
...certainly to date, focus on vocabs
19:00:56 [danbri]
timbl: for a purist, a model is just an ontology
19:01:09 [danbri]
pat hayes: did you find any exisitng ontology work of utilty, eg Cyc
19:01:14 [danbri]
kathy: did look at some of that stuff
19:01:22 [JosD___]
JosD___ has joined #swarch
19:01:28 [danbri]
randy: when i looked at daml.org, number of other vocabs was somewhat overwhelming
19:01:35 [danbri]
pat: not just daml, larger vocabs
19:01:39 [danbri]
danc: hard to see forest for trees!
19:01:43 [danbri]
randy: yup
19:02:05 [danbri]
timbl: if you met in street, you might say 'look at these three!' not list all of them
19:02:13 [danbri]
randy: didn't have any foaf in place ;-)
19:02:23 [danbri]
...we had eric to talk to, but not early on enough(?)
19:02:34 [danbri]
em: instead of going via me, could talk to partners
19:02:46 [danbri]
...not just paring things down
19:02:51 [DanC]
I attempted to do a 'look at these 3' in http://www.w3.org/RDF/#projects
19:03:01 [danbri]
...oftentimes you want to look at these thru a filter
19:03:07 [danbri]
randy: need metadata about the ontologies
19:03:21 [danbri]
kathy: i'd like to be able to say 'here are the 3 most commonly used'
19:03:30 [danbri]
pat: don't re-invent time and processes, they've been done
19:03:32 [danbri]
@@url!
19:03:42 [sandro]
Pat is suggesting: http://ontology.teknowledge.com/
19:03:43 [danbri]
timbl: the way you divide product and ocmpany up might be v sun-specific
19:03:47 [danbri]
thanks sandro :)
19:04:12 [danbri]
timbl: for those, modelling is yr own business, for other areas, more value in sharing
19:04:21 [danbri]
kathy: 90%ish we can share
19:04:32 [danbri]
ericp: any commitment to figuring out which that 90% is
19:04:36 [danbri]
randy: thats why we are here
19:04:41 [danbri]
kathy: yes, definitely
19:04:52 [danbri]
ian: re ontology, what tools are you using to build and maintain it
19:04:59 [danbri]
sandeep: custom built, based on java
19:05:06 [danbri]
...uses daml validator
19:05:13 [danbri]
...can validate(?) data with it
19:05:16 [danbri]
randy: pretty manual
19:05:27 [danbri]
...primary ontologist edits file in vi, then checks with validator
19:05:33 [danbri]
s: yes atthis point in time, vi
19:05:41 [danbri]
...in june timeframe, going for UI and more automation
19:05:46 [PStickler]
Ahhh, so they're doing it the right way ;-)
19:06:36 [danbri]
(scribe missed a few pts pluggin in laptop)
19:06:46 [danbri]
kathy: governance model, best practices...
19:06:50 [danbri]
... could learn from you guys
19:06:58 [danbri]
...internal, external... getting agreement is challenging
19:07:10 [danbri]
...onto best practice similarly
19:07:18 [danbri]
(next slide: technology opportunities)
19:07:58 [danbri]
- rdf aware search engines; improved navigation utilizing metadata; automated application of metadata; rdf/daml/owl-aware onto management tools; knowledge mininig and inference tools/approaches; topic maps leverage
19:08:10 [danbri]
randy: re knowledge mining
19:08:20 [danbri]
...this is where data gets v interesting, good to see something in this space
19:08:28 [danbri]
...also need to understand interaction, overlap with Topic Maps
19:08:43 [danbri]
...don't know where momentum from Knowledge Tech discussions re rdf/tm went
19:08:50 [danbri]
kathy: want to learn about vocabs to leverage
19:09:10 [danbri]
aside...ooOO(I wonder if RSS 1.0 might be useful for Swordfish... --danbri)
19:09:22 [danbri]
slide: critical enablers
19:09:27 [danbri]
...strong exec sponsorship is a big need
19:09:36 [danbri]
...herding cats, need for strong support
19:10:00 [danbri]
...education of the business re value proposition, not easy to get acrsos the 'whats in it for me'
19:10:12 [danbri]
...have to keep doing this, focus on practical savings
19:10:20 [danbri]
...business and technical partnership is key to success
19:10:27 [danbri]
...methodology: Sun Sigma based
19:10:40 [danbri]
...bus. process improvements, not just from a tech standpoint
19:10:50 [danbri]
...randy and i went out and sold this product almost as if a software product
19:10:52 [danbri]
thanks jos
19:11:23 [danbri]
...importance of relnship with groups such as this
19:11:26 [danbri]
...to see what is coming
19:11:31 [danbri]
...also to talk to other companies
19:11:42 [danbri]
...we are here with an interest in a Best Practice Sharing Working Group
19:12:09 [danbri]
contacts: program manager, kathy.macdougall@sun.com, chief architect: randy.willard@sun.com
19:12:13 [danbri]
frank: is this operational?
19:12:14 [danbri]
k: yes
19:12:21 [danbri]
randy: working its way into wider use
19:12:27 [danbri]
k: a web search that uses it
19:12:32 [danbri]
...its an infrastructure component
19:12:35 [danbri]
...that defines the stds
19:12:38 [danbri]
...not a website
19:13:02 [danbri]
frank: one issue, to what extent is this in operational use
19:13:06 [danbri]
kathy: it is
19:13:15 [danbri]
randy: yes, prod'n operation in real use
19:13:25 [danbri]
danc: re 30k triples thats just the ontology, right?
19:13:36 [danbri]
melissa: 3 mill assets being described...
19:13:47 [danbri]
kathy: also have data warehouse folk getting in touch
19:13:59 [danbri]
...re-using vocab to connect structured and unstructured data
19:14:06 [danbri]
randy: swordfish is just metadata registry piece
19:14:26 [danbri]
(re sigma, i found http://www.sun.com/2000-1115/sigma/ for more context)
19:14:42 [danbri]
ralphs: re going out and teaching folk this world's lingo, do they often have their own models...
19:14:50 [danbri]
...can you say more about issues you've encountered there
19:14:55 [danbri]
k: 1st thing def the lingo
19:15:01 [danbri]
...all of us agreeing on certain words for things
19:15:09 [danbri]
...i rarely tell people about rdf unless they are engineers
19:15:15 [danbri]
...focussing on what it is going to do for them
19:15:26 [danbri]
...if you do this to your content, the following things become more possible
19:15:32 [danbri]
randy: q was about content models?
19:15:40 [danbri]
ralph: yes... re unstructured vs structured...
19:15:58 [danbri]
... do you find when you go to a content group, they often have models that are useful to you when building an ontology
19:16:04 [danbri]
randy: infrequently...
19:16:40 [danbri]
k: helping educatiion of content owners
19:16:50 [danbri]
...some don't have an agreed content model
19:17:04 [danbri]
timbl: can you partition those you've talked to, eg. RDBMS vs Java vs ...
19:17:17 [danbri]
randy: sun has one of each of every kind of content management system
19:17:21 [danbri]
...often backened by oracel
19:17:27 [danbri]
...lots of way to carve things up
19:17:34 [danbri]
melissa: most content not in rdbms
19:17:36 [danbri]
...most is html
19:17:47 [danbri]
randy: one problem swordfish should solve...
19:17:54 [danbri]
...we have 100s of websites that support teams use
19:18:14 [danbri]
...this should help us rationalise and aggregate things
19:18:32 [danbri]
k: one thing from content agg standpoint
19:18:44 [danbri]
...we are NOT saying that we want one content management system across the company
19:19:03 [danbri]
...instead we say 'if you just use this common content model, common metadata, you can do what you want!
19:20:08 [danbri]
em: want to reduce barriers...
19:20:17 [danbri]
randy: like snowball just picking up on way downhill
19:20:24 [danbri]
...just coming out of first phase
19:20:40 [danbri]
harold: makes sense to build on top of an opensource content management system
19:20:45 [danbri]
eg plone.org (zope based)
19:20:49 [RalphS]
Tlone
19:20:50 [danbri]
...portal building, intranet
19:20:53 [danbri]
thanks ralph
19:21:46 [danbri]
timbl: ...when you've got it syntactically into rdf, you still have the prob of agreeing the common vocab within rdf's structure
19:21:48 [PStickler]
http://plone.org
19:22:09 [danbri]
...not just use cases, but supporting business of agreeing the 80% that can share a model
19:22:19 [danbri]
k: yes, business<->tech partnership is key
19:22:23 [danbri]
...need both sides on board
19:22:52 [RalphS]
thanks, Patrick -- I did hear wrong
19:22:53 [danbri]
kathy: express regrets that Bernard (Sun Labs) couldn't attend...
19:23:03 [danbri]
...having him take some requirements, keep dialog going
19:23:15 [danbri]
pat hayes: roughly how deep is your class hieraarchy
19:23:35 [danbri]
sandeep/randy: about three levels of class hierarchy
19:24:51 [danbri]
danbri: asked about whether looked into RSS 1.0
19:24:57 [danbri]
a: yes, early days though
19:25:04 [danbri]
harold: re rdf search engine?
19:25:16 [danbri]
...isn't every rdf query system a semantic search service
19:25:22 [danbri]
...you can always click through
19:25:33 [danbri]
...could get a hit list of links etc
19:25:54 [danbri]
kathy: looking for search engines, in traditional vein, that don't make use of embedded metadata
19:26:10 [danbri]
rrandy: rdf in html discussion
19:26:33 [danbri]
em: TAP and sematnic search seems relevance, augmenting of trad search w/ rdf-based additions
19:26:41 [danbri]
...more focussed areas
19:27:16 [PStickler]
SW search engines could crawl the SW via MGET just as existing Web engines crawl the Web using GET
19:27:25 [danbri]
ben: in daml program, there are various tools for such boosting
19:27:44 [danbri]
...one way is to do the inferecne to generate more terms, so simpler tools find the doc.
19:28:23 [danbri]
danbri: works across the web: google finds html pages that are generated from rdf descriptions
19:28:33 [danbri]
timbl: interested re discovery, browsing...
19:28:55 [danbri]
...normally if you go to a site, you often go to support section, or downloads section
19:29:04 [danbri]
...versus lookup by serial number
19:29:11 [danbri]
...evenentually you get the item not a page
19:29:21 [danbri]
...you can identify the specific operating system, driver, etc
19:29:29 [danbri]
...after that, i can plunge back into html world
19:29:50 [danbri]
...things liek the foafnaut, where you have a known item lookup, i find very nice
19:30:01 [danbri]
...if you know exactly which item to lookup, navigation can be very crisp
19:30:12 [PStickler]
Sounds like TimBL will like browsing the SW with MGET
19:30:21 [danbri]
...i'd like to see much more powerful rdf browsers, starting at an item and folding out linked things
19:30:39 [danbri]
...that blows a spreadsheet out of the water in terms of analytic facilities, interaative exploration
19:31:04 [danbri]
...in a very defined environment, browsing works well
19:31:29 [danbri]
timbl: it was just scripting plus SVG
19:32:05 [PStickler]
A generic RDF "browser" requires machinery such as MGET to achieve consistent behavior across SW enabled servers...
19:32:39 [sandro]
Or machinery such as Tim's recommended uses of # :-)
19:33:18 [chaalsBOS]
Guus Schreiber. Some thoughts on Semantic Web Best practices
19:34:24 [PStickler]
Using the # approach and arbitrary RDF documents is like doing a GET and getting a dozen HTML pages in response. What is needed is a well defined concept of 'concise bounded resource description' which forms the basis for consistent behavior of SW servers
19:35:06 [chaalsBOS]
GS: a goal is to help people set up their first semantic web app.
19:35:49 [chaalsBOS]
... so there are some vocabularies available, such as wordnet (in several versions) which we use in our own applications.
19:36:34 [danbri]
I have a wordnet rep too, and agree a common one would be useful. http://lists.w3.org/Archives/Public/www-rdf-interest/1999Dec/0002.html http://xmlns.com/2001/08/wordnet/
19:36:59 [libby]
http://swordfish.rdfweb.org/discovery/2003/02/stats/foafclasses.jsp?schema=http://xmlns.com/wordnet/1.6/ :)
19:37:18 [RalphS]
VRA - Visual Resources Association; specialization of Dublin Core
19:37:19 [RalphS]
http://www.vraweb.org/
19:37:35 [RalphS]
sigh, appears to require flash
19:37:36 [chaalsBOS]
... GS shows his VRA vocabulary - a specialisation of Dublin Core that can be used as a model for people to copy
19:38:16 [chaalsBOS]
... There are a lot of things out there that can be used, so it isn't always necesary to construct something from scratch.
19:38:53 [chaalsBOS]
... It doesn't take (experts at least) very long to convert a vocabulary to an RDF format, but making them available would be very helpful.
19:39:38 [chaalsBOS]
... Which raises the issue of how to deal with the different ontologies/vocabularies.
19:39:59 [Nobu]
Nobu has joined #swarch
19:40:02 [chaalsBOS]
Mik Dean: Should there be people creating data about whether they recommend a particular ontology?
19:40:14 [chaalsBOS]
GS There are people working on these areas
19:40:20 [chaalsBOS]
... it is a bit subjective
19:40:32 [chaalsBOS]
Pat Hayes: There are some tools that support this
19:41:06 [libby]
nice idea danc
19:41:11 [chaalsBOS]
Ian Horrocks: There are methodologies developed for describing characteristics of ontologies as metrics for quality.
19:42:41 [chaalsBOS]
GS It is easier to develop applications and publish them in public or semi-public domains (e.g. medicine, etc)
19:42:49 [danbri]
interesitng point
19:43:12 [chaalsBOS]
GS 2: Guidelines, FAQs, etc:
19:43:43 [chaalsBOS]
It would be good to have an archive of guidelines and examples for various common tasks
19:44:36 [chaalsBOS]
... transforming a vocabulary into RDFS/OWL, mapping ontologies, integration of various information sources,
19:44:51 [chaalsBOS]
... combining RDF with XML/HTML, etc...
19:46:14 [chaalsBOS]
GS 3: Tools and demos. Ones that show you using the tools as your own production systems are effective demonstrations
19:47:26 [Nobu]
Nobu has joined #swarch
19:47:28 [chaalsBOS]
... things that have nice pictures are good demos.
19:48:16 [chaalsBOS]
... (and it is important to have a real system behind them - "live demos" not special easy cases)
19:49:00 [chaalsBOS]
GS 4: Links to related techniques
19:49:59 [chaalsBOS]
GS Publishing technical notes describing relationships between similar approaches is useful - the relationship between Semantic Web and Topic Maps, etc...
19:50:30 [chaalsBOS]
Kathy: A lot of this is the stuff that we have done or wanted to do
19:50:38 [chaalsBOS]
Benjamin: Ditto
19:52:17 [chaalsBOS]
Benjamin: What is our story about how to work with people who are thinking in XML (presumably other than RDF/XML) - topic for open discussion.
19:53:00 [chaalsBOS]
Liddy Nevile gives presentation.
19:54:07 [chaalsBOS]
LN: Looking at how to talk to people about why replace a system that works with something new - what gets people asking about RDF in the first place.
19:54:34 [chaalsBOS]
... "making the magic explicit" - getting people to see some reason why RDF is cool.
19:55:49 [chaalsBOS]
... I spent a long time working on Logo because I saw some cool magic in Turtle Graphics that could lead to kids doing lts of cool stuff
19:56:25 [danbri]
liddy: 'knowledgeable is being good at knowing'
19:56:27 [danbri]
liddy++
19:56:29 [chaalsBOS]
... We all talked about the turtle as a way to do cool things, but some people didn't see the magic.
19:56:46 [danbri]
(using 'knowledge' as a mass noun doesn't really work, k as skill much better...)
19:58:16 [chaalsBOS]
LN: There are lots of rough diagrams, but there are very few well-developed demonstrations.
19:58:59 [chaalsBOS]
... I started to use FOAF in an aboriginal community.
19:59:26 [chaalsBOS]
... Problem is that the society model is very different, so "normal" relationships are mostly meaningless.
20:00:11 [chaalsBOS]
... Showing that FOAF could be readily used for a completely different group of relationship types was something that showed the value in that scenario.
20:00:20 [danbri]
I guess you can't just tell 'em "go make yr own namespace..."
20:00:30 [danbri]
(that works on rdf geeks...)
20:01:51 [chaalsBOS]
LN: It is important to explicitly show the thing the makes us excited - going one level past where we get the A-Ha, to show what it was we see that is so cool.
20:01:59 [chaalsBOS]
Break time...
20:02:18 [DanC]
======== break 'till 15:15
20:02:18 [RalphS]
http://www.googlism.com
20:25:36 [libby]
libby has joined #swarch
20:26:04 [Nobu]
Nobu has joined #swarch
20:33:38 [DanC]
[... discussion of syntax, quantification, ... ]
20:34:38 [DanC]
PatH: I think the approach of DL in RDF could be extended to all of FOL. Ugly, but seems doable.
20:36:43 [danbri]
would be intresting to see the details...
20:45:26 [DanC]
[... discussion of whether RDF(S) entailment is actually implemented, and hence whether there's something else that should be the basis of future work...]
20:45:51 [GuusS]
GuusS has joined #swarch
20:47:00 [sandro]
TimBL: Ben's saying people should using owl:Class unless they really mean rdfs:Class.
20:51:54 [sandro]
Controversy over how bad it is for systems to silently assume rdfs:Class means owl:Class.
20:52:19 [sandro]
Peter: you wont be able to see the difference if it's owl-DL.
20:57:11 [DanC]
ISWC in Oct
20:57:16 [DanC]
WWW2003 in May
20:57:24 [DanC]
DAML meeting in April
20:57:38 [DanC]
^possible places to meet
21:00:21 [danbri]
adjourned======================================
21:00:52 [danbri]
ACTION: ericm to send out slides to RDF IG lists, soonish.
21:11:22 [danbri]
danbri has left #swarch
21:19:35 [Nobu]
Nobu has joined #swarch
21:29:20 [Nobu]
Nobu has joined #swarch
21:53:18 [libby]
libby has joined #swarch