08:19:16 RRSAgent has joined #dwbp 08:19:16 logging to http://www.w3.org/2014/04/01-dwbp-irc 08:19:24 rrsagent, make logs public 08:19:27 Today seems here our Eduroam connection is down 08:19:30 * o 08:19:46 meeting: DWBP WG F2F Day 2 08:19:51 chair: Hadley 08:19:52 Phil how do you do those starry comments? 08:20:55 Is zzzzakim awake yet for phone service? 08:20:57 I can't dial in to Zakim 08:20:59 jeremy has joined #DWBP 08:21:25 Ig_Bittencourt_ has joined #DWBP 08:21:45 agenda: https://www.w3.org/2013/dwbp/wiki/London_2014 08:22:01 deirdrelee has joined #dwbp 08:22:02 would be nice to have hangout today too 08:22:50 markharrison has joined #dwbp 08:23:05 hangout setup, will post the link in a second... 08:23:17 "the conference is restricted at this time" 08:23:29 HadleyBeeman has joined #dwbp 08:23:30 yaso has joined #dwbp 08:23:31 zakim, room for 4? 08:23:33 ok, PhilA; conference Team_(dwbp)08:23Z scheduled with code 3927 (DWBP) for 60 minutes until 0923Z 08:23:38 zakim, code? 08:23:38 the conference code is 3927 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), PhilA 08:23:59 new google hangout for this morning here https://www.w3.org/2013/dwbp/wiki/London_2014#Google_Hangout 08:24:02 Team_(dwbp)08:23Z has now started 08:24:09 +ericstephan 08:24:26 fkyanai has joined #dwbp 08:24:29 antoine has joined #dwbp 08:24:45 +laufer 08:25:11 HadleyBeeman: I want us to split into group and work on use cases and vocabularies 08:25:18 ... we don't have much for each of them 08:25:35 ... we can start by creating wiki pages like we did yesterday for best practices 08:25:45 https://www.w3.org/2013/dwbp/wiki/Best_practices_notes 08:26:01 Good morning Makx, we're in a different room today 08:26:08 BernadetteLoscio has joined #dwbp 08:26:15 https://www.w3.org/2013/dwbp/wiki/London_2014#Preparation 08:26:20 ... it would be great if we had other pages and can take headers from documents that are online 08:26:34 ... for best practices we can steal headers from mobile best practices 08:26:37 ... etc 08:27:00 -ericstephan 08:27:01 Team_(dwbp)08:23Z has ended 08:27:01 Attendees were ericstephan 08:27:12 ... each group will need a scribe and to keep track of actions coming out of discussions 08:27:38 gatemezi_ has joined #dwbp 08:27:53 Team_(dwbp)08:23Z has now started 08:28:00 +Steve 08:28:03 +ericstephan 08:28:05 JoaoPauloAlmeida_ has joined #dwbp 08:28:44 ... we can look at use cases and existing work that is relevant 08:29:05 +[IPcaller] 08:29:12 Zakim, IPcaller is me 08:29:12 +JoaoPauloAlmeida_; got it 08:29:14 zakim, Steve has Hadley, Yaso, Flavio, Newton, Bernadette, Ig, Deirdre, Jeremy, Adriano, Brian, Laufer, Vagner, JohnG, MarkH, Antoine, Caroline, PhilA 08:29:14 +Hadley, Yaso, Flavio, Newton, Bernadette, Ig, Deirdre, Jeremy, Adriano, Brian, Laufer, Vagner, JohnG, MarkH, Antoine, Caroline, PhilA; got it 08:29:16 deirdrelee: we can split challenges into the three groups 08:29:48 Good morning! 08:29:53 good morning 08:30:38 Zakim, IPcaller is me 08:30:38 sorry, JoaoPauloAlmeida, I do not recognize a party named 'IPcaller' 08:30:56 I am having some technical problems, sound is very bad 08:30:57 Challenges google doc https://docs.google.com/spreadsheet/ccc?key=0AhTZf3B9yQ3odGVvU3pBazFsY3pyUVppNDFSZGtyQkE&usp=sharing&richtext=true#gid=5 08:31:22 not great sound at this point.... 08:31:31 Sound was good yesterday, but quite bad today 08:31:33 might be part of a dream sequence 08:32:13 on second tab, master challenges 08:32:13 It might be too tough with the breakouts 08:32:14 sorry, joaopaulo and ericstephan. We're trying to make it better — it's a bigger room. 08:32:16 the sound for me is a little low 08:32:35 is the facility under water? 08:32:44 sound on hangout is reasonable 08:32:51 what's different from yesterday? 08:33:01 joaopaulo: it's a different room 08:33:01 same problem here .. 08:33:21 ok 08:33:21 IRC might be our best friend today 08:33:32 better than yesterday, i guess it's room acoustics 08:33:36 I'm afraid there's not much we can do. I think ericstephan is right: IRC may be even more important than usual. 08:33:43 I'll try the sound in hangout 08:34:12 deirdrelee: the first one is metadata and it relates to vocabularies as well 08:34:21 adrianov has joined #dwbp 08:34:25 ... I propose metadata would go in best practices group 08:34:54 -ericstephan 08:34:57 ... the second challenge is the granularity of data to publish 08:35:20 ... formats - CSV and other formats 08:35:32 ... different indictors 08:35:42 ... heterogeneous formats 08:35:55 I've updated the agenda for today: https://www.w3.org/2013/dwbp/wiki/London_2014#2nd_day_.2801.2F04.29 08:35:59 ... common vobs are not used 08:36:14 ... added value comes from comparable data sets combined 08:36:29 ... usw 08:37:27 zakim, who is here? 08:37:27 On the phone I see Steve, JoaoPauloAlmeida_ 08:37:29 Steve has Hadley, Yaso, Flavio, Newton, Bernadette, Ig, Deirdre, Jeremy, Adriano, Brian, Laufer, Vagner, JohnG, MarkH, Antoine, Caroline, PhilA 08:37:29 On IRC I see adrianov, JoaoPauloAlmeida, gatemezi_, BernadetteLoscio, antoine, fkyanai, yaso, HadleyBeeman, markharrison, deirdrelee, Ig_Bittencourt_, jeremy, RRSAgent, Zakim, 08:37:29 ... PhilA, Caroline_, JohnGoodwin, adler1, laufer, newton, ericstephan, nathalia, gatemezi, MakxDekkers, trackbot 08:37:37 -JoaoPauloAlmeida_ 08:37:42 present+ Carlos 08:37:58 PhilA, again this code is not valid 08:38:13 I am just hanging out on IRC 08:40:33 +MakxDekkers 08:40:56 adler1, please do not forget to scribe on behalf of us remote participants 08:41:04 zakim, steve has me 08:41:04 +HadleyBeeman; got it 08:42:04 +[IPcaller] 08:42:14 Zakim, IPcaller is me 08:42:14 +JoaoPauloAlmeida; got it 08:43:33 deidrelee is reading all the challenges from the spreadsheet 08:43:45 +ericstephan 08:43:47 Ig_Bittencourt has joined #DWBP 08:44:30 I'm not hearing nothing 08:45:01 Hangout is mute 08:45:06 no sound at all nathalia? 08:46:10 no sound at all 08:46:18 at Hangout 08:47:41 No more parties can be added this time 08:47:46 :( 08:47:53 Zakim said this 08:47:55 zakim, room for 1 more? 08:47:55 I don't understand your question, PhilA. 08:48:04 zakim, room for 2? 08:48:06 sorry, PhilA; could not schedule an adhoc conference; passcode overlap; if you do not have a fixed code you may try again 08:49:10 deirdrelee is almost at the end of the challenges spreadsheet 08:49:47 Okay I'll hang up and wait on IRC PhilA 08:49:48 Zakim, who is talking? 08:49:50 ok PhilA 08:49:54 CarlosIglesias has joined #dwbp 08:49:54 -ericstephan 08:50:00 JoaoPauloAlmeida, listening for 10 seconds I heard sound from the following: MakxDekkers (31%), Steve (13%) 08:50:01 Zakim, who is making noise? 08:50:13 newton, listening for 10 seconds I heard sound from the following: MakxDekkers (57%), Steve (86%) 08:50:18 MakxDekkers, can you mute yourself? 08:50:27 zakim, mute me 08:50:27 MakxDekkers should now be muted 08:50:45 Is it the rain in Spain Makx? 08:50:57 +[IPcaller] 08:51:18 I'm at Zakim 08:51:27 ... they could all go into best practice and pull out the ones into data usage and granulatiry and they others can go into the other group 08:51:38 zakim, ipcaller is nathalia 08:51:38 +nathalia; got it 08:51:44 :)) 08:52:15 q+ 08:52:20 had: our next step is to plit into groups 08:52:57 s/plit/split/ 08:53:25 s/had:/HadleyBeeman 08:53:44 I could be in group 1 if I good hear :-O 08:54:07 which are the groups? 08:54:20 Thing 1 is best Practices 08:54:25 had: the discussion groups will discuss use cases and the structure of the documents 08:54:28 Thing 2 is Quality & Granularity 08:54:32 Thing 3 is Data Usage 08:54:34 creating requiremenst from the challenges 08:54:50 the requirements we come up with will go into the use case docuemnt 08:54:59 Not writing the vocabs, but working on requirements for those requirements 08:55:20 who are the discussion leaders - each group decides 08:56:03 ok 08:58:35 Can't we merge Thing 2 and Thing 3 ? 09:02:54 gatemezi_: the decisions may merge, but first we need to sort out the notes and use cases for each. 09:03:27 Ok, HadleyBeeman 09:03:32 sound is gone worse now 09:03:38 During this breakout time I'm wondering for productivity if it would be helpful to spend some time doing analysis between syergistic/overlapping CSVW Use Cases and DWBP Challenges. Would this be too divergent? 09:04:04 leigh dods is doing some work on licensing in ODI and we agreed to park that idea and come back to it later 09:04:26 I was suggesting that we split off licences, i.e. machine readable licences. Invite Leigh Dodds, and perhaps the ODRL folks, to look at that. This WG *could* be a vehicle for standardisation of that work if a) this WG felt it useful; b) extra capacity came to the WG to deliver it 09:05:03 action: PhilA to contact Leigh Dodds wrt exchangeable licences etc. 09:05:03 Created ACTION-17 - Contact leigh dodds wrt exchangeable licences etc. [on Phil Archer - due 2014-04-08]. 09:05:55 for the record i am also to do some work on licenses for europeans 09:06:14 so, what's going on? 09:06:51 HadleyBeeman, what's going on? 09:06:52 JoaoPaulo, Makx, Eric: 09:06:53 I'm thinking is dividing groups 09:07:05 (and nathalia, gatemezi) 09:07:29 It's so crazy that video feed is good and sound terrible... 09:07:33 We are splitting up into 3 groups. One for Best practices, one for Data usage vocab, one for data quality vocab. 09:07:46 Give us a sec — I will tell you how to join 09:07:51 thx 09:07:52 ok 09:07:56 ok 09:07:58 Ok great, thank you 09:08:24 fkyanai has joined #dwbp 09:09:46 yaso has joined #dwbp 09:10:38 newton_ has joined #dwbp 09:12:36 -MakxDekkers 09:12:46 çlaufer 09:12:55 -laufer 09:13:04 Best Practices groups is working on IRC #dwbpbestpractices 09:13:13 laufer has left #dwbp 09:13:26 dwbp-DataUsage 09:13:42 Data Usage Task force: #dwbp-DataUsage 09:14:10 -nathalia 09:16:13 Okay, remote participants: we are splitting into the groups, each with their own IRC channel. 09:16:19 newton_ has left #dwbp 09:16:31 BrianMatthews has joined #DWBP 09:16:44 ok, What about sound/hang out? 09:16:47 They are working on notes on the wiki: https://www.w3.org/2013/dwbp/wiki/Main_Page#Deliverables 09:17:14 IRC Channel for quality and granularity group is #dwbpq&g 09:17:33 JoaoPauloAlmeida, there is a google hangout in the Best Practices discussion, and a Skype connection in the Data Quality discussion 09:18:16 -JoaoPauloAlmeida 09:18:48 And a hangout for the Data Usage discussion too, I think. 09:19:17 MakxDekkers has joined #dwbp 09:19:20 https://www.w3.org/2013/dwbp/wiki/Data_quality_notes 09:19:52 Zakim, who's here? 09:19:52 On the phone I see Steve 09:19:54 Steve has HadleyBeeman 09:19:54 On IRC I see MakxDekkers, BrianMatthews, yaso, fkyanai, CarlosIglesias, adrianov, JoaoPauloAlmeida, BernadetteLoscio, antoine, HadleyBeeman, markharrison, deirdrelee, jeremy, 09:19:54 ... RRSAgent, Zakim, PhilA, Caroline_, JohnGoodwin, adler1, nathalia, gatemezi, trackbot 09:20:43 MakxDekkers has joined #dwbp 09:21:04 PhilA, when you publish a dataset its for the potential users of the dataset is this a one off and will it be updated every month 09:21:26 ... what ever the update period is where is the policy that says you are going to do so 09:22:06 topic: Data quality task force 09:23:03 .. DCAT is missing things that we can add to 09:23:36 q+ 09:24:13 ack PhilA 09:24:17 ack MakxDekkers 09:25:00 MakxDekkers: the subjective quality parameters are more difficult to capture and represent 09:25:18 ... should be careful not to mix these with the objective aspects, which will lead to quality 09:25:49 adler1: could we include both? 09:25:50 PhilA: how to express something that is subjective 09:27:08 jeremy, metrics are subjective to each person point of view 09:27:26 BrianMatthews has left #dwbp 09:27:34 ... we encounter a number of subjective metrics 09:28:00 disconnecting the lone participant, Steve, in Team_(dwbp)08:23Z 09:28:01 Team_(dwbp)08:23Z has ended 09:28:01 Attendees were ericstephan, JoaoPauloAlmeida_, Hadley, Yaso, Flavio, Newton, Bernadette, Ig, Deirdre, Jeremy, Adriano, Brian, Laufer, Vagner, JohnG, MarkH, Antoine, Caroline, 09:28:01 ... PhilA, MakxDekkers, HadleyBeeman, JoaoPauloAlmeida, nathalia 09:30:27 BrianMatthews_ has joined #DWBP 09:31:08 adler1: quality should be bidirectional, user opinions should be incorporated 09:31:49 can we separate and start with the objective facts? 09:32:09 PhilA: there is a w3c workshop on this in san francisco 09:32:24 -> http://www.w3.org/2014/04/annotation/ Annotations workshop 09:32:45 a subjective statement is not a fact about a resource 09:34:06 it's a fact about the relationship between the person that makes a jusdgment and the resource 09:35:09 zakim, end meeting 09:35:09 I don't understand 'end meeting', PhilA 09:35:16 zakim, bye 09:35:16 Zakim has left #dwbp 09:35:25 Zakim has joined #dwbp 09:35:33 zakim, room for 6? 09:35:34 ok, PhilA; conference Team_(dwbp)09:35Z scheduled with code 3927 (DWBP) for 60 minutes until 1035Z 09:35:41 those are objective facts deirdre 09:35:52 yes 09:35:54 zakim, code? 09:35:54 the conference code is 3927 (tel:+1.617.761.6200 sip:zakim@voip.w3.org), PhilA 09:36:17 +q 09:36:32 Team_(dwbp)09:35Z has now started 09:36:38 +Steve 09:37:17 jeremy: different domains have different quality requirements 09:37:41 ack MakxDekkers 09:38:26 MakxDekkers, the provider will never say its rubbish whereas the user may provide that feedback 09:38:29 antoine: many quality metrics can be added by data publisher or user, it's data centric 09:39:23 MakxDekkers: the data publisher and user will add different kinds of metrics, we should make a distinction between statement of facts and opinions 09:43:08 antoine: why close the door on data user adding objective facts? 09:44:24 deirdrelee: we can first look at obective facts and then subjective, but who adds the objective facts (users or publishers) is irrelevnat 09:45:27 adler1: user is not only an individual, but could also be an institution. publisher could also be an individual 09:46:05 antoine: for example OSM 09:47:58 -> http://www.w3.org/2014/04/daQ Jeremy's slides 09:53:00 jeremy: a colleague of Jeremy's in Uni of Leipzig did a paper on quality metrics, should be available soon 09:53:30 does daQ include information about the entity that did the measurement? 09:55:29 jeremy: the provenance should be included in vocab, but at what level, per metric, per dimension? 09:56:23 PhilA: there could be multiple metrics on a particular dataset 09:57:05 antoine: our role would be to define what metrics to include? 09:57:51 PhilA: jeremy has to convince dwbp chairs that we should use daQ 09:58:15 phi 09:59:28 PhilA: Evaluation and Report Language (EARL) W3C accessibility has some of what jeremy can reuse http://www.w3.org/WAI/intro/earl.php 09:59:52 ... it includes provenance etc. 10:00:23 jeremy: nobody whould question the data in the quality graph 10:00:47 antoine: why not? 10:01:05 zakim, who is here? 10:01:05 On the phone I see Steve 10:01:07 On IRC I see BrianMatthews_, MakxDekkers, yaso, fkyanai, CarlosIglesias, adrianov, JoaoPauloAlmeida, HadleyBeeman, markharrison, deirdrelee, jeremy, RRSAgent, PhilA, Caroline_, 10:01:07 ... JohnGoodwin, adler1, nathalia, trackbot 10:01:46 antoine has joined #dwbp 10:02:37 deirdrelee: jeremy's model is for automatically created metrics, not necessarily human-created 10:03:56 PhilA: are there liability issues with quality metrics? 10:04:21 adler1: there could be multple values for each metric 10:04:39 Looking at EARL schema. Looks quite useful 10:04:44 http://www.w3.org/TR/EARL10-Schema/ 10:05:29 PhilA: can jeremy add a real use-case? 10:06:45 ... any metric that you measure is a subset of those that could be measured, the selection of metrics is subjective 10:07:25 adler1: quality can be a composite point of view 10:07:37 antoine: we will still have to define some metrics 10:07:52 PhilA: jeremy is doing this 10:08:33 antoine: if there are multiple use-cases that point to same metrics, we can include them in the vocab, but they can be extended 10:10:26 PhilA: there are some datasets that are authoritative, this should be a metric, but is this a quality metric or a descriptive metric (for dcat+) 10:12:29 jeremy: we should think about what consumers need to know 10:13:07 deirdrelee: data publishers could also say their data is not good quality 10:13:21 antoine: what is the relationship with datacube? 10:14:08 adler1: there are metrics for confidence, but no metric for measuring doubt, 10:14:33 ... to allow human nature to admit some doubt 10:15:03 deirdre a publisher would most likely not say unqualified "not good", they would probably say things like ... 10:15:28 this data hasn't bee ncleaned, or it comes from a source we haven't been able to verify 10:15:38 ... don't only have positive assertiions on data, how about negative assertions, but phrased in a way to make it 'pschologically' ok for data not to be great 10:15:51 that's what I meant 10:16:06 the negative or disclaimers 10:16:27 cheers makx 10:16:45 chris taggart from opencorporates chatting to group now 10:17:26 can't hear chris at all 10:17:50 he's just chatting, not a presentation 10:18:34 we're having a coffee break now too 10:18:39 chat in 10 mins 10:19:40 so is it sensible to say there are three levels: 1. dcat+-type facts 2. daQ/EARL-type metrics 3. opinions? 10:19:43 This is the survey paper I was talking about: www.semantic-web-journal.net/system/files/swj556.pdf - it is still under review, but a good overview of what metrics exists for LOD 10:21:40 useful refernce. 10:38:29 Maybe also useful http://www.slideshare.net/OpenDataSupport/open-data-quality-29248578 10:39:19 Slide 8 has an overview of quality dimensions 10:39:24 cheers 10:40:00 disconnecting the lone participant, Steve, in Team_(dwbp)09:35Z 10:40:02 Team_(dwbp)09:35Z has ended 10:40:02 Attendees were Steve 10:40:44 @MakxDekkers - did you actually implement those metrics? 10:41:46 jeremy, these are not metrics -- the slides provide best practice advice to publishers to address those dimensions 10:41:57 we haven't measured the data they provide 10:43:42 yes you are right... my mistake.. they would be "Dimensions" in terms of daq - (as you rightly said in the slides) 10:43:52 BP recomm is eg "check data before publication to increase consistency" (slide 17) 10:47:31 has the group reconvened yet? 10:49:22 I hear the word "quality" but I can't hear what people are saying. 10:49:41 no,everyone's still out for coffee 10:49:50 ok 10:49:51 steve and i talking to chris 10:59:41 ericstephan_ has joined #dwbp 11:06:30 fkyanai has joined #dwbp 11:11:24 yaso has joined #dwbp 11:13:52 the issue of trust is a completely different aspect 11:14:12 the model needs to allow for lying 11:14:44 reputation is certainly an issue 11:14:51 Scribe: JoaoPauloAlmeida 11:14:53 sorry! 11:15:03 that would be complicated 11:15:08 Scribe: JohnGoodwin 11:15:38 any statement about quality needs to be accompanied by provenance info 11:16:10 dcat+-type statements are covered by the fact that it is the publisher saying it 11:16:37 and you either trust the publisher or you don't 11:17:23 deirdre, yes let's get concrete 11:19:17 sound is confused 11:21:12 several people talking 11:25:20 adler1: reputation is important