Semantic Web Interest Group IRC Chat Logs for 2007-05-04

00:29:14 <danja_> damn. why am i even tryingto catch up with stuff. insert old connolly quote here

00:30:30 * danja_ blabbers on #swhack

02:13:52 * dajobe scrolls back a day

02:16:33 * dajobe scratches head at http://www.getsemantic.com/wiki/Triplr_operator_userscript

02:58:53 <jar286_> jar286_ is now known as jar286

10:46:41 <JimJibber> JimJibber is now known as JibberJim

12:25:28 <jar286_> jar286_ is now known as jar286

13:33:23 <chimezie>http://triplr.org/turtle/http://planetatom.net/

13:33:26 <dc_swig> A: http://triplr.org/turtle/http://planetatom.net/ from chimezie

13:33:42 <chimezie> A:|Extracting AtomOWL triples from aggregated atom feeds

13:33:43 <dc_swig> Titled item A.

13:35:10 <chimezie> A:From a single aggregated atom feed the following modalities are produced: XHTML+RDFa (for human consumption), [http://planetatom.net/exhibit|JSON] (for use with Exhibit), and RDF via content negotiation and GRDDL

13:35:12 <dc_swig> Added comment A1.

13:35:55 <chimezie> A1:From a single aggregated atom [http://planetatom.net/feed/|feed] the following modalities are produced: XHTML+RDFa (for human consumption), [http://planetatom.net/exhibit|JSON] (for use with Exhibit), and RDF via content negotiation and GRDDL

13:35:57 <dc_swig> Replaced comment A1.

13:36:10 <chimezie> A:|Extracting AtomOWL triples from aggregated atom feeds (Planetatom.net)

13:36:12 <dc_swig> Titled item A.

13:37:08 <chimezie> A:->swig,atom,grddl,rdfa,xhtml

13:37:10 <dc_swig> Set keywords for A.

13:37:51 <chimezie> A:The Exhibit modality uses the timeline widget as a way to sift through the feed visually

13:37:53 <dc_swig> Added comment A2.

13:38:15 <chimezie> A:With faceted browsing support keyed off atom categories and authors

13:38:16 <dc_swig> Added comment A3.

13:39:10 <chimezie> A:A SPARQL endpoint is forthcoming..

13:39:11 <dc_swig> Added comment A4.

13:40:28 <dumnut> hi

13:41:42 <chimezie> A1:From a single aggregated atom [http://planetatom.net/feed/|feed] the following modalities are produced: XHTML+RDFa (for human consumption), [http://planetatom.net/output/index.js|JSON] (for use with Exhibit), and RDF via content negotiation and GRDDL

13:41:43 <dc_swig> Replaced comment A1.

13:42:02 <chimezie> A:The Exhibit [http://planetatom.net/exhibit|view] uses the timeline widget as a way to sift through the feed visually

13:42:03 <dc_swig> Added comment A5.

13:42:12 <chimezie> A2:The Exhibit [http://planetatom.net/exhibit|view] uses the timeline widget as a way to sift through the feed visually

13:42:14 <dc_swig> Replaced comment A2.

13:42:14 <chimezie> A5:

13:42:16 <dc_swig> (chimezie) The Exhibit [http://planetatom.net/exhibit|view] uses the timeline widget as a way to sift through the feed visually

13:42:25 <chimezie> A5:""

13:42:27 <dc_swig> Deleted comment A5.

13:43:33 <chimezie> A1:From a single aggregated atom [http://planetatom.net/feed/|feed] the following modalities are produced: [http://planetatom.net/|XHTML+RDFa] (for human consumption), [http://planetatom.net/output/index.js|JSON] (for use with Exhibit), and RDF via content negotiation and GRDDL

13:43:35 <dc_swig> Replaced comment A1.

13:46:31 <kidehen> chimezie: see: [http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html?uri=http%3A%2F%2Fplanetatom.net%2Ffeed%2F]

13:46:43 <kidehen> then go to the TimeLine

13:47:07 <chimezie> okay

13:49:56 <chimezie> i see the categories (on the far right), so it looks like it loaded it.. but I don't see a link to TimeLine

13:56:47 * chimezie wins with http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html?uri=http%3A%2F%2Fplanetatom.net

13:57:09 <chimezie> i wonder if it got the Atom OWL triples by conneg or GRDDL

14:01:28 <chimezie> kidehen, is there a good whitepaper on openlink virtuoso?

14:29:22 * timbl wondering what to say about linked data on dev day in banf

14:30:02 <timbl> Anyone in Chris Bizer's group around?

14:35:46 <njero_> njero_ is now known as njero

14:37:53 <timbl> I think it would be nice to have a good list of linked data sources in the wiki

14:39:47 <ericP> i'm fuzzy on what makes data "linked data"

14:39:53 <LeeF> me too!

14:40:18 <ericP> perhaps the hcls data counts <inquisitive shrug/>

14:43:46 <timbl> Well, lucky I wrote an article on it then!

14:44:14 <timbl>http://www.w3.org/DesignIssues/LinkedData

14:44:15 <dc_swig> B: http://www.w3.org/DesignIssues/LinkedData from timbl

14:44:32 <timbl> B:|Linked Data -- what makes data linked data

14:44:33 <dc_swig> Titled item B.

14:45:38 <timbl> B: 1. Use URIs as names for things

14:45:38 <timbl> B: 2. Use HTTP URIs so that people can look up those names.

14:45:38 <timbl> B: 3. When someone looks up a URI, provide useful information.

14:45:40 <timbl> B: 4. Include links to other URIs. so that they can discover more things.

14:45:41 <timbl> There.

14:45:43 <dc_swig> Added comment B1.

14:45:43 <dc_swig> Added comment B2.

14:45:43 <dc_swig> Added comment B3.

14:45:43 <dc_swig> Added comment B4.

14:46:09 <timbl> Stuff you can rampage about in wit the tabulator etc

14:47:08 <timbl> Like dbpedia

14:49:09 <kjetilk> kidehen works on dbpedia, but just left

14:49:31 <tommorris> Some nicely Linked Data: http://tools.opiumfield.com/twitter/tommorris/rdf

14:51:48 <kidehen> chimezie: There are collections of white papers re. Virtuoso :-)

14:52:01 <kjetilk> logger: pointer

14:52:01 <kjetilk> See http://chatlogs.planetrdf.com/swig/2007-05-04#T14-52-01

14:52:18 <kjetilk> kidehen: you might want to have a look at the backlog while you were away

14:52:20 <ericP> timbl, i'd skimmed the DI, but it mostly seemed like the standard SW good practice

14:52:39 <kidehen> kjetilk: ok :-)

14:53:02 <ericP> perhaps i missed an icremental step

14:53:14 <timbl> standard SW good rpactice ... wish it wrre more common

14:55:02 <ericP> as do i.

14:55:03 <ericP> just trying to understand the branding here

14:56:00 <timbl> For example, if a seminar series ha a set of seminars, thne make a URI for the series as well as each seminar, and link them, so one can link to he series.

14:57:09 <timbl> hmmm i need a starting pijn t in tommorris/rdf

14:57:41 <ericP> interesting. instead of { ( :one :two :three) a :SeriesList . }

14:58:53 <kidehen> kjetilk: are you referring to <http://tools.opiumfield.com/twitter/tommorris/rdf> via http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html or http://dbpedia.org ?

14:59:34 <kjetilk> dbpedia, or more generally timbl's request for linked data in the wiki

14:59:40 <timbl> ericp - the list would work too.

15:00:30 <timbl> tommorris, for example, if I load that file, I can start at http://twitter.com/CaptSolo

15:00:34 <ericP> timbl, but it doesn't give you a GETtable ID for the list; just a bnode

15:01:15 <tommorris> Just cracking open my editor to change that.

15:01:15 <ericP> and it's that GETtable resource that's the goal of "linked data" ?

15:01:37 <timbl> yesm you need a gettable ID

15:05:13 <kidehen> timbl: I am thinking that we should somehow conjure up a "Linked Data" specific white paper in tutorial? Incorporate your notes, some demos and screencasts etc..

15:05:48 <kidehen> timbl: or do you think the session for next Friday will cover this in enough detail?

15:07:09 <kidehen> timbl: I see comprehension of the "Give everything a URI" message the crux of the matter across many target audiences. I think we need to make this a pivot point re. Web and Web in the Enterprise messaging etc..

15:08:12 <kidehen> kidehen: URIs (from my vantage point) deliver: Data, Information, and Knowledge (depending on users context) at our finger tips

15:08:55 <kidehen> and the Linked Data best practice is a great demonstration of this powerful reality

15:14:02 <tommorris> Okay, another SemWebber's Twitter RDF: http://tools.opiumfield.com/twitter/danja/rdf

15:14:29 <tommorris> I've added a 'primaryTopic' element and 'rdfs:seeAlso'.

15:15:18 <kidehen> try: http://demo.openlinksw.com/DAV/home/demo/Public/Queries/DataWeb/CaptSolo_Twitter.isparql (this is both a dynamic data page and a linked data demo in one)

15:16:35 <kidehen> it uses content negotiation to determine it's behavior so you can also input into <http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html> or Tabulator as a dereferencable URI (with links to other URIs)

15:16:41 <tommorris> That's coming from the RSS/Atom, I presume?

15:17:29 <kidehen> in the current incarnation its on the fly transformation of non RSS 1.0 feeds (RSS 2.0 or Atom ) into RSS 1.0 :-)

15:17:49 <tommorris> i'm just XSLing the API.

15:18:24 <kidehen> yes since Virtuoso has an XSLT processor inside

15:20:55 <kidehen> tommorris: here is a version of the dynamic data page using your data: http://demo.openlinksw.com/DAV/home/demo/Public/Queries/DataWeb/TomMorris_Twitter.isparql

15:22:36 <kidehen> tommorris: scary URL, but here is a permalink from the RDF browser combining your Twitter with CaptSolo's:

15:22:37 <kidehen> http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html?uri[]=http%3A%2F%2Fdemo.openlinksw.com%2FDAV%2Fhome%2Fdemo%2FPublic%2FQueries%2FDataWeb%2FCaptSolo_Twitter.isparql&uri[]=http%3A%2F%2Fdemo.openlinksw.com%2FDAV%2Fhome%2Fdemo%2FPublic%2FQueries%2FDataWeb%2FTomMorris_Twitter.isparql&

15:22:39 <dc_swig> C: http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html?uri[]=http%3A%2F%2Fdemo.openlinksw.com%2FDAV%2Fhome%2Fdemo%2FPublic%2FQueries%2FDataWeb%2FCaptSolo_Twitter.isparql&uri[]=http%3A%2F%2Fdemo.openlinksw.com%2FDAV%2Fhome%2Fdemo%2FPublic%2FQueries%2FDataWeb%2FTomMorris_Twitter.isparql& from kidehen

15:23:14 <kidehen> but it fails :-(

15:23:30 <tommorris_> Yes, the other day I put together a SPARQL query to show me a list of friends in common between my profile and a friends

15:23:35 <timbl> kidehen, you still have to make the 'query' button press itself on those direct URIs

15:23:51 <kidehen> timbl: oops! Yes

15:24:43 <kidehen> tommorris: so try: http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html?uri=http%3A%2F%2Fdemo.openlinksw.com%2FDAV%2Fhome%2Fdemo%2FPublic%2FQueries%2FDataWeb%2FTomMorris_Twitter.isparql

15:24:56 <kidehen> and then add http://demo.openlinksw.com/DAV/home/demo/Public/Queries/DataWeb/TomMorris_Twitter.isparql

15:25:08 <kidehen> then click on the TimeLine tab

15:25:54 <timbl> I must say the stuff inside the openlink browser seems really good but the UI is not something i owuld hand to a grandmother

15:26:06 <timbl> """_:519_1

15:26:06 <timbl> null:http://twitter.com/Tolchock

15:26:07 <timbl> name:Maki"""

15:26:09 <timbl> ??

15:26:32 <tommorris_> My mum is now using Zotero... :)

15:27:02 <timbl> Is she a grandmother?

15:27:06 <kidehen> timbl: Zitgist is for grandmom :-)

15:27:15 <tommorris_> No, she's not a grandmother.

15:27:24 <timbl> ah well the ;-)

15:27:28 <kidehen> timbl: http://zitgist.com

15:27:28 <timbl> then

15:27:33 <tommorris_> But she does have an RDF database on her desktop to store her research data.

15:28:08 <kidehen> timbl: I think the count down can safely commence in about 3 weeks time re. Grandma and the Data Web :-)

15:28:29 <kidehen> timbl: lots of Linked Data etc.. re. People, Places, and Other things

15:28:43 <timbl> What's coming - something sepcific or just the general momentum?

15:29:09 <kidehen> timbl: a service that plugs grandma into the semantic data web :-)

15:29:23 <dmwaters> {global notice} Good day all. We had a problem with one of our rotation servers that segfaulted. We are currently working on the problem, and i will give any further updates in wallops, thank you for your time, and thank you for using freenode!

15:30:01 <kidehen> timbl: this is what we've been working on with fgiasson for a while now

15:30:41 <kidehen> timbl: we have grandmas as early adopters :-)

15:32:32 <timbl> ok, I'm all for it.

15:33:39 <timbl> tomorris, is there any id in that file which i could save as a starting point for browsing?

15:34:23 <timbl> All the URIs are in http://twitter.com/ its seems

15:43:10 <tommorris_> I've edited the file so that the RDF has a root element that points to the main 'profile' document

15:44:48 <tommorris_> I'm not sure whether I can put in <rdf:Description rdf:about="twitter.com/..." and rdf:ID="tom">

15:52:32 <DanC_lap> linked data resources in the wiki... yeah, timbl, I've had that same feeling a few times lately

15:53:10 <DanC_lap> I found http://esw.w3.org/topic/NetworkedData , which is odd in that it seems similar but I can't tell if it's causally related

15:55:35 <DanC_lap> C:|something with a URI so long it messes up the formatting of swig.xmlhack.com

15:55:37 <dc_swig> Titled item C.

15:59:18 <kidehen> Danc: When you say: Wiki which platform?

15:59:46 <kidehen> DanC: http://esw.w3.org ?

16:00:18 <kidehen> DanC: We already have this done for MediaWiki and we do have MoinMoin scheduled

16:01:01 <kidehen> DanC: This is via RDF Views of the SQL data in these platforms

16:01:17 <kidehen> DanC: same applies to phpBB and WordPress

16:01:21 <DanC_lap> B:see also [http://esw.w3.org/topic/LinkedData|LinkedData] in the esw wiki

16:01:23 <dc_swig> Added comment B5.

16:01:42 <LeeF> Might this be what people are looking for? http://esw.w3.org/topic/TaskForces/CommunityProjects/LinkingOpenData/DataSets

16:02:15 <DanC_lap> seems related, but that's clearly owned by some taskforce.

16:02:22 <timbl> tomorris: no you can'thave about and id

16:02:41 <LeeF> right. ok, so more like - this seems useful to create what you're looking for :)

16:02:42 <timbl> Use owl same as to link #tom with your twitter id

16:03:27 <DanC_lap> sorta ironic to put CommunityProjects *under* something

16:03:44 <DanC_lap> seems to me that the space for CommunityProjects is http://esw.w3.org/topic/

16:04:07 <timbl> Leef - that is i think just what i want

16:04:19 <DanC_lap> yes, it has the right content

16:04:25 <timbl> exept i'd like a starting URI for a *thing* in each one

16:04:34 <timbl> an example place to start browsing

16:04:37 <DanC_lap> I added a link from http://esw.w3.org/topic/LinkedData

16:04:50 <timbl> The name is kinda long ... i thought wikis were flat

16:05:05 <DanC_lap> wikis were flat. until somebody added hierarchy. :-/

16:05:13 <tommorris_> timbl: I've added the following triple: <#> foaf:PrimaryTopic <http://twitter.com/{username}>.

16:05:31 <DanC_lap> P or p ?

16:05:32 <timbl> heirarchies were trees, until someone added links ;-)

16:05:40 <tommorris_> sorry, p

16:05:57 <DanC_lap> and doesn't the person go on the pointy end of primaryTopic ?

16:06:03 <timbl> <#> should be i think

16:06:06 <DanC_lap> i.e. <page> foaf:primaryTopic <#person>.

16:06:10 <timbl> sorry should be <>

16:06:41 <DanC_lap> or is <> a page about a page?

16:06:47 <tommorris_> The RDF/XML is: <rdf:Description rdf:about=""><foaf:primaryTopic rdf:resource="http://twitter.com/danja"/></rdf:Description>

16:06:48 <timbl> <> primary topic <<http://twitter.com/{username}> works

16:07:43 <tommorris_> I thought about doing it as pure FOAF, but there are numerous non-human users on Twitter.

16:07:49 <DanC_lap> "Currently being RDFized by a W3C Best Practices Task Force" hmm... still current?

16:08:17 <DanC_lap> whether the user is a person or not, surely <http://twitter.com/{username}> is a page, not a peson?

16:08:28 <tommorris_> Yep

16:08:44 <DanC_lap> so your RDF/XML is RDF about a page?

16:09:02 <tommorris_> Well, the data is coming from an API.

16:09:11 <LeeF> logger, pointer?

16:09:11 <LeeF> See http://chatlogs.planetrdf.com/swig/2007-05-04#T16-09-11

16:09:45 <timbl> hmmm i have cleared my Firefox cache ... did yo change the version on the web, tommorris?

16:09:53 <timbl> i don't see the linl

16:09:54 <tommorris_> Yep

16:10:35 <tommorris_> Try it with another username: ldodds, iand, keithalexander etc

16:10:57 <timbl> I see it in cwm .. probably a cache problem

16:11:18 <tommorris_> The date format will change when I'm not snowed in with work

16:12:08 <kidehen> DanC: added links to the Wiki age

16:12:15 <kidehen> s/age/page

16:13:13 <timbl> content-type: text/html

16:14:42 <tommorris_> Fixed. I use text/html so I can just use my browser to debug it

16:15:54 <timbl> Ahhhh! works now.

16:16:18 <timbl> 85 acquaintainces

16:17:30 <tommorris_> I'll be adding 'followers' and any other data I can squeeze from their API and/or website.

16:17:49 <timbl> Lots of people have names, though you don't is af far as i can see

16:18:53 <tommorris_> yes, that's the API call - which only returns the friends list. The other data will get merged in soon enough

16:20:56 * DanC_lap is curious to see how GRDDL news spreads... finds enough stuff at http://www.technorati.com/search/grddl?language=en&authority=n to threaten my other todo items today...

16:25:03 <timbl> Hmmm the dataset list is intruiguing but I'm not having much luc k

16:25:15 <timbl> has http://esw.w3.org/topic/WordNet been hacked I wonder

16:26:39 <DanC_lap> grddl news takes me to Elias's blog which takes me to the WWW2007 dev track... wow. how am I going to get any hiking in? http://www2007.org/prog-Developers.php

16:28:27 <DanC_lap> B:note the linked data session [http://www2007.org/prog-Developers.php#friday|Friday in the WWW2007 Dev track]

16:28:29 <dc_swig> Added comment B6.

16:30:11 * DanC_lap updates http://esw.w3.org/topic/LinkedData to note that session

16:31:30 <chimezie> A:You can mouse-over the embedded RDFa with ([http://www.w3.org/2001/sw/BestPractices/HTML/rdfa-bookmarklet/|RDFa-bookmarklets] - in firefox)

16:31:32 <dc_swig> Added comment A5.

16:35:21 <timbl> danc, did you just write http://esw.w3.org/topic/LinkedData or a while ago?

16:35:31 <DanC_lap> just started it

16:36:32 <DanC_lap> I noodled about starting it a couple days ago (http://swig.xmlhack.com/2007/05/01/2007-05-01.html#1178032640.286146 )

16:38:51 <DanC_lap> say, kidehen , have you run the GRDDL tests against virtuoso? care to share the results?

16:39:05 * DanC_lap is reading a rather lengthy review of virtuoso http://www.mkbergman.com/?p=355

16:39:24 <kidehen> DanC: yes

16:39:37 <kidehen> DanC: Will put together a page and expose a URI :-)

16:39:45 <DanC_lap> spif.

16:39:51 <kidehen> Chimezie: didn't you do this a while back

16:40:19 <DanC_lap>http://www.w3.org/TR/grddl-tests/

16:40:21 <dc_swig> D: http://www.w3.org/TR/grddl-tests/ from DanC_lap

16:40:28 <DanC_lap> D:|GRDDL Test Cases

16:40:29 <dc_swig> Titled item D.

16:40:35 <kidehen> DanC: may re-read: http://www.openlinksw.com/blog/~kidehen/?id=1144 (Hello Data Web...)

16:40:35 <DanC_lap> D:W3C Working Draft 2 May 2007

16:40:37 <dc_swig> Added comment D1.

16:41:04 <timbl> How do you link to something down in a directory structure in wiki?

16:41:07 <DanC_lap> D: last call! Comments are due by 31 May 2007 to public-grddl-comments@w3.org

16:41:08 <dc_swig> Added comment D2.

16:41:08 <kidehen> DanC: will add live demos link to the GRDDL implementations page

16:41:31 <DanC_lap> timbl, ["Something/Horrible/Like/This"] . not very wiki, IMO

16:42:02 <DanC_lap> one wonders why not <a href="something/standards/like/this"> at that point

16:42:28 <timbl> aye, one doesn, one does...

16:43:41 <DanC_lap> the tension between a-space-for-our-group and a-space-we-share-with-the-world is tangible in the esw wiki

16:44:46 <timbl> maybe stuff can get promoted with alink

16:44:50 <DanC_lap> D:best sort of comment is [http://www.w3.org/TR/grddl-tests/#earl-reporting an implementation report in EARL/RDF]

16:44:51 <dc_swig> Added comment D3.

16:45:08 <timbl> But taking the forst one, I can't find a wordnet word in rdf

16:46:05 <DanC_lap> yeah, I'd like the list grouped by "exhibits all 4 linked data principles", "3 stars; good try", "2 stars, getting there" and "1 star, at least you stuck your stuff in the web"

16:46:55 <DanC_lap> maybe 0.5 stars for the .zip files

16:47:10 <DanC_lap> at least the .zip files are currently separated

16:49:11 <RalphS> TimBL, see http://www.w3.org/2006/03/wn/wn20/instances/word-bank

16:49:44 <RalphS> err, in your case, http://www.w3.org/2006/03/wn/wn20/instances/word-bppl

16:49:48 <RalphS> err, in your case, http://www.w3.org/2006/03/wn/wn20/instances/word-book

16:50:38 <timbl> Ok, so they are both words.

16:50:45 <RalphS>http://www.w3.org/2006/03/wn/wn20/instances/wordsense-book-noun-1

16:50:46 <timbl> There are no links to other words?

16:50:46 <dc_swig> E: http://www.w3.org/2006/03/wn/wn20/instances/wordsense-book-noun-1 from RalphS

16:51:33 <RalphS>http://www.w3.org/2006/03/wn/wn20/instances/synset-book-noun-1

16:51:34 <dc_swig> F: http://www.w3.org/2006/03/wn/wn20/instances/synset-book-noun-1 from RalphS

16:51:53 <RalphS> wn20schema:hyponymOf

16:52:44 <RalphS>http://www.w3.org/TR/2006/WD-wordnet-rdf-20060619/#intrown

16:52:45 <dc_swig> G: http://www.w3.org/TR/2006/WD-wordnet-rdf-20060619/#intrown from RalphS

16:52:48 <timbl> I can't find a relationship between 'read' and 'read' (past and future)

16:52:57 <timbl> s/future/present

16:52:58 <RalphS> Figure 1. The class hierarchy of the WordNet schema

16:53:42 <RalphS>http://www.w3.org/2006/03/wn/wn20/instances/synset-read-verb-1

16:53:43 <dc_swig> H: http://www.w3.org/2006/03/wn/wn20/instances/synset-read-verb-1 from RalphS

16:54:22 <timbl> Ah!

16:54:38 <timbl> Ok, let us use that as an example in th elist of datasets

16:59:04 <chimezie> oh wow I didn't quite know how to use 'live' the wordnet 2.0 naming convention

16:59:32 <kidehen> DanC: http://esw.w3.org/topic/LinkedData re. Live Demos section

16:59:57 <kidehen> I've added a Browser session permalink and a dynamic data page uri

17:02:43 * timbl fixes spam in http://esw.w3.org/topic/WordNet

17:30:20 <bigjb__> bigjb__ is now known as bigjb_

17:41:06 <timbl> I think there is a bug in tabulator when you have lots of sameAs .. it doens't look them all up

17:50:23 <dmwaters> {global notice} Hi all! freenode is currently looking for both main rotation servers and a hub in Europe. if you are interested in sponsoring a server for freenode please see: http://freenode.net/hosting_ircd.shtml Thank you for your patience, and thank you for using freenode!

18:01:22 <c> hrm

18:01:38 <c> techmeme needs clustering to avoid 90% of the stories being about YHOO

18:01:55 <c> is that deri store in a pub facing SVN or such or just press-release overload?

18:12:20 <timbl> ?

18:16:34 <rhythm>http://slashdot.org/articles/07/05/04/1252215.shtml

18:16:35 <dc_swig> I: http://slashdot.org/articles/07/05/04/1252215.shtml from rhythm

18:17:06 <rhythm> A:| slashdot comments face off over large RDF store and general semantic-web feasability

18:17:08 <dc_swig> Titled item A.

18:17:53 <rhythm> woops

18:18:02 * rhythm no ida how to erase

18:20:28 <rhythm> A:|""

18:20:30 <dc_swig> Titled item A.

18:21:05 <rhythm> I:| useless Slashdot comments on DERI 7-gigaquad store

18:21:06 <dc_swig> Titled item I.

18:24:08 <sbp> A:|PlanetAtom on Triplr

18:24:10 <dc_swig> Titled item A.

18:24:57 <sbp> I:These days it's often better and more enlightening to read the tags on Slashdot entries than all the comments. Slashdot tags are probably the only decent use for tags anybody's ever found.

18:24:58 <dc_swig> Added comment I1.

18:30:13 <rhythm> thx. irssi bug keeps deleting channels so i didnt know what A: was :)

18:30:13 <rhythm> rhythm is now known as c

18:31:17 <sbp> could always check out the online logs

19:16:49 <kidehen> timbl: http://demo.openlinksw.com/DAV/JS/rdfbrowser/index.html now has "Query Execution" when you hit "Enter"

19:16:54 <kidehen> you can try: http://www.w3.org/TR/grddl-tests/

20:02:49 <HarryH> Chime - are you around?

20:05:01 <chimezie> im here

20:05:02 <chimezie> sorta

20:07:00 <HarryH> Ah, do you have enough time to walk through health-care example in primer quickly?

20:07:18 <chimezie> i think so

20:07:22 <HarryH> OK.

20:07:27 <chimezie> where is you're roadblock?

20:07:39 <chimezie> s/you're/your

20:08:17 <HarryH> Ah, the OWL bit.

20:11:37 <chimezie> okay, so an HL7 document is a communication between healthcare 'components' : machines, departments, etc..

20:12:27 <chimezie> the transform takes an instance of the an HL7 CDA document and generates an RDF graph which describes content using terms in this vocabulary: http://purl.org/cpr/0.5#

20:12:52 <HarryH> I see that.

20:13:26 <HarryH> It's this bit which is confusing me:

20:13:29 <chimezie> I can either continue this way, or you can let me know what specific part you want me to clarify

20:13:30 <chimezie> okay

20:13:43 <HarryH> ex:DiagnosingImage a owl:Class;

20:13:44 <HarryH> rdfs:subClassOf foaf:Image,

20:13:44 <HarryH> [ a owl:Restriction;

20:13:44 <HarryH> owl:onProperty ex:indicates;

20:13:44 <HarryH> owl:someValuesFrom cpr:medical-sign ].

20:13:46 <HarryH> OK.

20:14:19 <HarryH> So what you have is a subclass of an image with a restriction that states the property ex:indicates is a cpr:medical-sign?

20:14:27 <chimezie> okay so that creates a class of 'images' which are actually XRays that indicate a 'problem'

20:14:40 <HarryH> Yep.

20:15:17 <HarryH> However, that syntax is giving cwm and pellet issues.

20:15:34 <HarryH> As dajobe noted, there seems to be missing a rdfs:subClassOf somewhere.

20:15:42 <chimezie> what particular issues?

20:16:14 <HarryH> Just wait a sec...I think it just breaks cwm when I try to convert it over to RDF/XML.

20:16:15 * chimezie is not sure what is wrong with the OWL statements

20:16:26 <HarryH> What OWL reasoner do you use?

20:17:04 <chimezie> i use my own mostly (Fuxi).. not a tableux-based reasoner like pellet, but a rule based engine (like cwm)

20:17:14 <chimezie> for forward chaining, not consistency checking

20:17:29 <chimezie> that turtle expands to :

20:17:31 <HarryH> Ah, I thought I had fuxi installed somewhere...

20:17:35 <chimezie> ex:DiagnosingImage a owl:Class

20:17:47 <chimezie> ex:DiagnosingImage rdfs:subClassOf foaf:Image

20:17:55 <chimezie> ex:DiagnosingImage rdfs:subClassOf [ .. etc.. for restriction .. ]

20:18:48 <chimezie> some values creates an 'existential' class of things where there is an ex:indicates relationship between 'something' and a member of the cpr:medical-sign class

20:19:00 <myren__> myren__ is now known as myren_

20:19:45 <chimezie> the latter rdfs:subClassOf relationship says ex:DiagnosingImage is a member of this class - i.e., all ex:DiagnosingImages have an ex:indicates relationship with some 'medical sign'

20:20:15 <HarryH> When I run cwm I get:

20:20:17 <HarryH> swap.notation3.BadSyntax: Line 14 of <http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.owl>: Bad syntax (newline found in string literal) at ^ in:

20:20:17 <HarryH> "... rdfs:comment "Property relating a foaf:Image to a medical ^

20:20:17 <HarryH> sign it indicates";

20:20:17 <HarryH> rdfs:subPropertyOf [ owl:i..."

20:20:31 <HarryH> Wait a sec..that's an artifact..

20:20:35 <HarryH> let me fix that and try again..

20:20:38 <chimezie> okay

20:20:55 * chimezie finds a stand alone snippet to parse with n3p

20:21:13 <HarryH> I'm using:

20:21:14 <HarryH>http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.owl

20:21:15 <dc_swig> J: http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.owl from HarryH

20:21:26 <chimezie> ahh thx

20:21:30 * chimezie tries lsw

20:21:59 <chimezie> (which triggers pellet)

20:22:06 <HarryH> OK.

20:22:08 <HarryH> Got it working.

20:22:14 <HarryH> What is lsw?

20:22:48 <chimezie> it's an OWL visualizing tool by alan ruttenberg - hidden away in a google group (i'm not sure why he doesn't have a home page for it)

20:22:50 * chimezie finds a link

20:23:13 <HarryH> OK, now I need to figure out the pellet command to identify the bit of the RDF we need.

20:23:27 <HarryH> We have to demo this to people at WWW2007 who have never used OWL before likely.

20:23:29 <chimezie> i see a pellet error

20:23:49 <HarryH> We could have people in the tutorial install fuxi or something else.

20:23:52 <chimezie> well lsw ouputs taxonomy diagrams

20:23:58 <chimezie> nice eye candy

20:24:04 <HarryH> That could be cool for the tutorial.

20:24:09 <chimezie>http://groups.google.com/group/lsw-discuss/browse_thread/thread/3d6f1c3d46ac77bb

20:24:11 <dc_swig> K: http://groups.google.com/group/lsw-discuss/browse_thread/thread/3d6f1c3d46ac77bb from chimezie

20:24:22 <chimezie> K:Alan R. on how to export screenshots from LSW

20:24:24 <dc_swig> Added comment K1.

20:24:50 <chimezie> K:[http://groups.google.com/group/lsw-discuss/browse_thread/thread/5006f92addb5868a|lsw 0.9 release]

20:24:52 <dc_swig> Added comment K2.

20:25:03 * chimezie goes back to pellet error

20:25:46 <chimezie> hehe.. lsw works with rdf/xml not n3 - I'll have to re-serialize that

20:26:11 <HarryH> Already did it..

20:26:23 <chimezie> link?

20:26:37 <HarryH> Checking it in..

20:26:39 <chimezie> okay

20:27:03 * chimezie gets http://rafb.net/p/S2CwRp66.html

20:28:14 <HarryH>http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.rdf

20:28:15 <dc_swig> L: http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.rdf from HarryH

20:29:00 <chimezie> pellet doesn't seem to have a problem with that particular snippet , let me try that link you just posted

20:29:15 <HarryH> I guess the question is pedagogical.

20:29:28 <HarryH> Should we have them run pellet on the snippet first?

20:30:12 <chimezie> oic, you want to be able to demonstrate the entailment as part of the demo not just talk about it?

20:30:22 <HarryH> Yep.

20:30:32 <HarryH> Ideally, I want the users in the room to:

20:30:36 <HarryH> 1) Install Pellet

20:30:46 <HarryH> 2) Run it over the RDF generated from the HL7 XML file.

20:30:49 <HarryH> (by GRDDL)

20:30:57 <HarryH> 3) Get the entailment results.

20:31:08 <HarryH> 4) Re-merge entailment results into original RDf

20:31:19 <HarryH> 5) Then SPARQL out the interesting bit, i.e. the diagnosis image.

20:31:24 <HarryH> I know it's a lot to ask.

20:31:31 <HarryH> :)

20:31:36 <chimezie> oohh.. that's right on the bleeding edge :)

20:31:59 <chimezie> how much of that does pellet give you for free?

20:32:16 <HarryH> I'm trying to figure it out, I don't use Pellet very often.

20:32:23 <HarryH> They obviously don't give us GRDDL.

20:32:28 <chimezie> i've only ever use it as a third party tool to SWOOP or LSW, never directly

20:32:46 <HarryH> How is SWOOP doing?

20:32:51 <HarryH> Could SWOOP or LSW help us here?

20:33:21 <HarryH> Ideally we want to remain java-based to keep things simple.

20:33:32 <HarryH> People in the tutorial are likely runnning windows :(

20:34:36 <chimezie> the seemed to move to google code but not completely - I've had trouble launching from their new code base

20:34:52 <chimezie> LSW is completely java - java with an embedded lisp processor

20:34:55 <HarryH> Argh.

20:35:05 <HarryH> Java with embedded Lisp processor?? That's wild.

20:35:13 <HarryH> I didn't think you could embed Lisp in Java.

20:35:17 <HarryH> But I guess I was wrong.

20:35:20 <chimezie> very wild

20:35:38 <chimezie> well..

20:36:28 <HarryH> Hmmm...pedagogically not sure how to go here.

20:36:39 <HarryH> Let me see if this isn't horrible from the pellet commandline.

20:36:41 <chimezie> you might get more of it across if you step through the entailment instead of trying to demo it with a hodge podge of tools which are still behind the state of the art (which is a moving target)

20:36:45 <HarryH> And I'll try installing lwm.

20:37:03 <HarryH> What do you mean "step through the entailment?"

20:37:11 <HarryH> Like walk through what the OWL reasoner is doing?

20:37:15 <chimezie> yes

20:38:04 <HarryH> Hmm...I could do that.

20:38:12 <chimezie> i think the only rule that applies is:

20:38:13 <chimezie> ?R owl:onProperty ?P; owl:someValuesFrom ?C. ?S ?P ?M. ?M a ?C} => {?S a ?R}.

20:39:20 <chimezie> {?P owl:inverseOf ?Q. ?S ?P ?O} => {?O ?Q ?S}.

20:39:40 <chimezie> {?P @has rdfs:domain ?C. ?S ?P ?O} => {?S a ?C}.

20:39:44 <chimezie> {?P @has rdfs:range ?C. ?S ?P ?O} => {?O a ?C}.

20:39:49 <chimezie> s/rule/rules

20:40:22 <chimezie> {?SP rdfs:subPropertyOf ?R. ?P rdfs:subPropertyOf ?SP} => {?P rdfs:subPropertyOf ?R}.

20:42:22 <HarryH> Still hanging out with Pellet's commandline.

20:42:51 <chimezie> you have the starting RDF graph, the owl rules, the new triples you entail, the merge, and a SPARQL query over the closure dataset

20:43:22 <HarryH> Yep.

20:43:31 <HarryH> The problem is really the merge.

20:43:54 <HarryH> I would just do it with cwm or rapper.

20:44:09 <HarryH> But...hmm...I guess I could see if Jena can pull it.

20:44:33 <HarryH> I'm just not trying to horrify the users :)

20:44:45 <HarryH> by making them install all sorts of things.

20:45:53 <HarryH> Do we have any Pellet commandline users in #swig?

20:46:03 * chimezie remembers an n3 merge diagram from a timbl slide that went down easy

20:46:27 <HarryH> Ah, if you can find that it would be great.

20:47:10 <chimezie> http://www.w3.org/2000/10/swap/Examples.html

20:47:35 <chimezie> color coding for asserted/entailed triples

20:54:33 <HarryH> Hmmm...

20:55:04 <HarryH> This example does seem to be driving Pellet for a loop for some reason.

20:55:26 <chimezie> oh you also need:

20:55:27 <chimezie> {?C rdfs:subClassOf ?SC. ?M a ?C} => {?M a ?SC}.

20:56:07 <HarryH> Ah, it's claiming it's not OWL-DL.

20:56:19 <chimezie> i wonder if it's the rdfs:subPropertyOf [ owl:inverseOf cpr:interpretant-of ];

20:57:04 <HarryH> It seems to be claiming it's just rdf:types being missing.

20:57:40 <chimezie> .. rdfs:subPropertyOf [ a owl:ObjectProperty; owl:inverseOf cpr:interpretant-of ]

20:57:48 <chimezie> s/a/rdf:type

20:58:10 <HarryH> Really? Why is that?

20:59:03 <chimezie> i've found Protege to complain about DL axioms used with untyped classes..

20:59:10 <chimezie> rather unpleasant struggle

20:59:24 <HarryH> I thought "a" was just rdf:type.

20:59:34 <chimezie> oh you mean the s/a/rdf:type

20:59:39 <HarryH> Yep.

20:59:39 <chimezie> yes..

20:59:48 <chimezie> it is

20:59:48 <HarryH> I thought that was just n3 for saying rdf:type.

20:59:53 <HarryH> Give me one second...

21:02:25 <chimezie> might also want to promote that blank node to a named owl:ObjectProperty

21:03:45 <HarryH> Here's what Pellet says:

21:03:46 <HarryH>http://www.w3.org/2001/sw/grddl-wg/doc29/pelletresults.owl

21:03:47 <chimezie> ex:indicates owl:inverseOf cpr:interpretant-of

21:03:48 <dc_swig> M: http://www.w3.org/2001/sw/grddl-wg/doc29/pelletresults.owl from HarryH

21:07:20 <chimezie> what is the input to that result?

21:10:32 <JimJibber> JimJibber is now known as JibberJim

21:47:54 <dmwaters> {global notice} hi all! I need to do some emergency maintenence on 2 rotation servers to help get more rotation servers back into the rotation. This will be mildly disruptive, and i will give more information in wallops as i go. /mode your_nick +w to see them. Thank you for your patience, and thank you for using freenode!

21:59:04 <njero_> njero_ is now known as njero

22:02:51 <jar286_> jar286_ is now known as jar286

22:10:02 <chimezie> HarryH: I took the same source document ()http://www.w3.org/2001/sw/grddl-wg/doc29/hl7-sample.rdf, modifed the diagnosing OWL snippet to http://rafb.net/p/MLK7SE86.html, and using a minimal ruleset (http://rafb.net/p/C54upX52.html) I was able to entail these triples: http://rafb.net/p/mPDIML83.html

22:10:28 <chimezie> notice:

22:10:30 <chimezie> [[

22:10:33 <chimezie> [ a _:blsOzgke192,

22:10:33 <chimezie> _25:DiagnosingImage;

22:10:33 <chimezie> _24:inAllOf _:blsOzgke191,

22:10:33 <chimezie> _:blsOzgke193;

22:10:33 <chimezie> _25:indicates [ a _:blsOzgke209,

22:10:34 <chimezie> cpr:medical-problem;

22:10:36 <chimezie> _24:inAllOf _:blsOzgke207]].

22:10:38 <chimezie> ]]

22:11:57 <chimezie> HarryH: the full closure was: http://rafb.net/p/eIDWP288.html

22:13:27 <chimezie> so you might want to replace http://www.w3.org/2001/sw/grddl-wg/doc29/diagnosingimage.rdf with http://rafb.net/p/MLK7SE86.html

22:16:01 <Md> [GlobalNotice] The IRC server you are connected to is being upgraded. In a few minutes you will be disconnected shortly.

22:16:25 <chimezie> logger, pointer?

22:16:25 <chimezie> See http://chatlogs.planetrdf.com/swig/2007-05-04#T22-16-25

22:17:26 <chimezie> phenny, tell HarryH to see http://chatlogs.planetrdf.com/swig/2007-05-04#T22-16-25 for how diagnosingimgae.rdf was modified to entail ex:DiagnosingImage statements (tested with Fuxi)

22:17:29 <phenny> chimezie: I'll pass that on when harryh is around.

22:18:15 Disconnected from irc.freenode.net (Connection reset by peer)

22:19:06 Topic now The topic for #swig is: Semantic Web Interest Group 24x7 hack-n-chat. logged for all to see. http://www.w3.org/2001/sw/interest/ http://swig.xmlhack.com/ * For the Interface Generator, try irc.oftc.net #swig

22:19:06 Users on #swig: logger Cloud Earle_Martin HarryH shellac_ yosemite dc_swig dajobe phenny deltab ericP das05r Emeka kasei pjenvey Yudai anselm jang drewp jipp jllykifsh workbench tag__ CaptSolo evlist alder terraces KjetilK_ nym er4z0r Ciantic dmiles_afk tuukkah sr jsvuorin Xavier nemequ LotR yvesr Pipian gjh dorian karamba38 myren_ sh3n dngor nwalsh ghard kpreid DanC_lap LeeF schepers iand kwijibo chimezie iwaim___ c kidehen bigjb_ Mutiny- sandro JibberJim

22:19:06 Users on #swig: sbp perigrin darobin bigjb shellac njero jar286 jaresty

22:19:06 <NickServ> Password accepted - you are now recognized

22:30:27 <chimezie> K: see [http://svn.mumble.net:8080/svn/lsw/tags/release-0.9/README.txt|README.txt]

22:30:28 <dc_swig> Added comment K3.

22:30:51 <dmwaters> {global notice} Good day folks. We're done with the emergency maintenence for the time being. We had this scheduled for 6am utc for the 5th but do to some dificulties it had to be pushed up a bit. I apologize for the inconvenience, and thank you for using freenode!

22:56:50 <sbp-> sbp- is now known as sbp

23:52:18 <jar286_> jar286_ is now known as jar286