SWXG Weekly Telecon

 

 

<hhalpin> 1. Convene SWXG WG meeting of 2009-08-19T13:00-15:00Z

<tinkster> (I think)

<hhalpin> chair: hhalpin

<tinkster> I don't mind.

<cperey> :-)

ok

<hhalpin> scribe: FabGandon

<hhalpin> PROPOSED: to approve SWXG WG Weekly -- 29th July 2009 as a true record

<hhalpin> http://lists.w3.org/Archives/Public/public-xg-socialweb/2009Aug/0006.html

<rreck> +1 approve minutes

<melvster> +1

<hhalpin> PROPOSED: to meet again Wed. Aug 26th

<cperey> +1

<hhalpin> RESOLVED: approved SWXG WG Weekly -- 29th July 2009 as a true record

<tinkster> +1

<rreck> +1 weekly meetings

<hhalpin> RESOLVED: meet again Wed. Aug 26th

hhalpin: we are back on weekly meetings.

<hhalpin> 2. General Organization and Task Forces

<hhalpin> ACTION: [DONE] Mischa to write up Soren and Jonathan's findings from last telecon for final report. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action01]

hhalpin: current time may be excluding people of west coast

<hhalpin> ACTION: [CONTINUES] DKA to summarize OSLO and geoLocation conversation in order to spread knowledge of these efforts among W3C members [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action02]

hhalpin: for the agenda of next meeting

<hhalpin> ACTION: [CONTINUES] hhalpin to summarize metrics and gypsii talk. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action03]

<hhalpin> ACTION: [DONE] danbri circulate wookie proposal to the xg [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action04]

<hhalpin> ACTION: [CONTINUES] Mischa to describe/implement a report of terms and conditions, and how they change between now and the end of the XG. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action05]

<hhalpin> ACTION: [CONTINUES] hhalpin to have http://www.w3.org/TR/vcard-rdf would refer to the new version, there will be a 'previous version' link to the current one [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action06]

<hhalpin> ACTION: [CONTINUES] hhalpin take action to clean up the web-page and make it less intimidating to those who aren't members [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action07]

<hhalpin> ACTION: [DONE] Melvster to draft columns=network rows=tech matrix for tech report [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action08]

danbri: wookie new apache project
... http://lists.w3.org/Archives/Public/public-xg-socialweb/2009Aug/0051.html
... "Wookie is a Java server application that allows you to upload and deploy widgets for your applications. Wookie is based on the W3C Widgets specification, but widgets can also be included that use extended APIs such as Google Wave Gadgets and OpenSocial."

<hhalpin> +1 to DanBri's wookie+ opera+widget proposal

<melvster> yep

<hhalpin> http://spreadsheets.google.com/pub?key=tL_pHzU2JRih72ULVgD0xuQ&output=html

hhalpin: melvin went through the use cases, deduced the functionalities,

<bblfish> yes

<tinkster> yes

melvster: put up a template for identifying what can be done with existing frameworks

<MacTed> Yes/Partial/No is good. "Planned" might be a good additional value?

hhalpin: so many possible use cases

<rreck> i think its helpful to keep the number, but each have an owner

<cperey> +1 with Ron's recommendation

<tinkster> 7.7 was my user story, so I can be made its champion.

hhalpin: some use cases are very technical other not, there are a lot of differences

<rreck> +1 common level of abstraction

hhalpin: should we use a questionnaire to compare them and get a feeling of what people think about

<cperey> the distinction is between the feature and the "how" it is implemented

<bblfish> but you don't want to make the use case/stories too complex

<rreck> feature is useful, i think we are not in a position to recommend the how

<hhalpin> quite tricky!

<bblfish> so you have to keep them simple

ted: you would need single sign on to to some oif these things

<rreck> we could say, i think it might be done using..<foo>

<hhalpin> (it might need to be in text).

ted: what have been implented that allows this use cases ?

<cperey> how the feature is implemented

<cperey> it's a third dimension?

henry: an additional problem is how there are implemented

<rreck> plus, as technology changes it would deprecate our how to analysis, whereas the feature isnt deprecated

henry: we may need an additional matrix / set of criteria

<hhalpin> aha

<Adam> to decompose the use case in to the parts that make it up?

<hhalpin> yes, I think we're heading into decomposition of use-case.

<bblfish> well it's all interesting

<rreck> i think #1 each gets a champion

<cperey> and a lot of work

<bblfish> there are some use cases that are way too big

<hhalpin> agreed bblfish.

hhalpin: instead of a questionnaire should we have 4 people looking at the use cases ?

<rreck> +1 champions call

<hhalpin> So, let's get champions.

<bblfish> ok, I am interested in this project.

<oshani> +1 for the questionnaire to select use cases

<hhalpin> does that make sense?

<bblfish> so we have micro groups for each use cases

<bblfish> makes sense

<melvster> yes, makes sense

<rreck> -1 dropping usecases that have champions

<bblfish> pair programming on use cases

hhalpin: vote on each use case and the one with enough champions will be kept and detailed.

<rreck> +1 drop duplicates

<hhalpin> ok...

ted: don't drop any of them unless they are duplicates.

<danbri> questionaire sounds useful yep

<rreck> or unless it doesnt have a champion

<bblfish> but we should simplify.

<hhalpin> ACTION: hhalpin to make WBS for use-cases champions [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action09]

<trackbot> Created ACTION-66 - Make WBS for use-cases champions [on Harry Halpin - due 2009-08-26].

<hhalpin> 3. User Stories

<cperey> remember that once we have been through the process, there could be maintenance

<hhalpin> ACTION: [CONTINUES] adam to write up the boeing use case for enterprise social networks [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action10]

<hhalpin> ACTION: [CONTINUES] hhalpin to write up the social network data exchange story [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action11]

<hhalpin> ACTION: [CONTINUES] tpa to look for similar use cases list to expand ours [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action12]

<hhalpin> ACTION: [CONTINUES] tinkster to add user story along lines of... if you enable import of contacts from other sites, then it lowers the barrier to entry for new people signing up at your site, hence get more users, hence get more advertising revenue. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action13]

<hhalpin> ACTION: [DONE] danbri to write up orkut/i18n/"looking for" issue [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action14]

danbri: other invited guesses

<hhalpin> ACTION: [CONTINUES] rreck to write up use case based on shill case [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action15]

hhalpin: open micro-bloging would be interesting to

<hhalpin> ACTION: [CONTINUES] rreck to send some thoughts on new template to be sent out to listserv. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action16]

<hhalpin> ACTION: [CONTINUES] danbri to find someone from Opera to talk Widgets [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action17]

<hhalpin> ACTION: [CONTINUES] danbri and karl to contact Evan of identi.ca over OpenMicroblogging and W3C. [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action18]

wiki page of invited guests : http://www.w3.org/2005/Incubator/socialweb/wiki/InvitedGuestSummaries

<hhalpin> ACTION: [CONTINUES] Renato and DanBri and hhalpin to schedule DataPortability.org/XMPP [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action19]

<hhalpin> 5. Guest Speaker: Nathan Eagle

<danbri> on my todo list then: xmpp, identi.ca, dataportability, widgets, microformats / ben ward

<hhalpin> 5. Guest Speaker: Nathan Eagle

<hhalpin> Nathan Eagle was brought to our attention by Hakan.

<hhalpin> algorithms trained on mobile phone date

nathan: http://web.media.mit.edu/~nathan/bio.html

hhalpin: how hard to get the data from the phones ?

<hhalpin> Reality-mining installed invisible spy phone on mobile phone

<hhalpin> and had to get informed consent.

Nathan: we used a logger starting on boot, logging everything, invisible for the user,
... hard to scale the installation of an app on every handset

<hhalpin> scale - the operators have all this data, can get this kinda data without installing spyware

Nathan: you need to have super computer to study the graph you obtain

<cperey> +q

<hajons> +q

<rreck> it would be neat if private privacy based companies had the data instead of the telecos

hhalpin:  we are also interested in the fact you have access in this sort of data

<rreck> thereby usurping telecos

<hhalpin> impossible to get the data without the telcos.

hhalpin: how hard for a mobile phone company to access these telco data ?

<rreck> im sure i could have the app send me the data

nathan: telco view these data as their property.
... these data are viewed more and more valuable and thus the grip is tightening

<rreck> an new OBAMA initiative!

cperey: the telco are also concerned about regulations.

nathan: telco are also scared of getting burned by the press
... only governments can force them to release the data.

<rreck> individuals should own their data we all agree but we have like few cards to play

nathan: the only use for this data right now is for juridical action

<rreck> if we vote with our dollar we would have no where to turn

<cperey> -q

hajons: different cases in different countries.
... hard to explain to the user what kind of data we need and why

<bblfish> I don't understand, on the iPhone every app can get access to geo location data

<rreck> imagine if you own the data and someone else holds it for you, you will still be at their mercy

hajons: very tricky area and regional problem.

<bblfish> so it should be real easy to create a service to post that info somewhere

nathan: PR is still a big deal in china, but users' reaction is different from one region to another

<rreck> laptop manufactures put chips in equipment that US govt doesnt like

<bblfish> CDR the format operators keep the telephone info in

<hhalpin> interested in this idea that a little app could be put on by installed that provides richer data, i.e. by handset manufacturers apple

<rreck> for instance Lenovo <sp> isnt allowed in certain places

nathan: a mobile phone company can install app hich produce much interesting data than the logs of operators.

<hhalpin> also interested in this fact that the location isn't in CDR unless a call is made...

hhalpin: did you deal with the reality of the variety of handsets

nathan: no we worked only with a set of donated nokia
... ? problem on the iPhone because of the background apps problem (?)

<tinkster> iphone background apps problem - apps running in the background effectively sleep.

<tinkster> ... can't do any processing.

<hhalpin> http://www.skyhookwireless.com/howitworks/

hhalpin: geolocation APIs, what needs to be standardized for contextual data in general?
... did you have access to contact books?

<hhalpin> http://www.pnas.org/content/early/2009/08/14/0900282106.abstract

nathan: we show how top link behaviour and real network of friends

<hhalpin> Full article: http://www.pnas.org/content/early/2009/08/14/0900282106.full.pdf+html

also: http://sciencenow.sciencemag.org/cgi/content/full/2009/817/1

<danbri> (v interesting)

extract: Just by analyzing the calling patterns, the researchers could accurately label two people as friends or nonfriends more than 95% of the time.

hhalpin: how to validate friendship

<rreck> i wonder if you could infer fellow terrorist cell members? i wonder who might fund that analysis.

nathan: no way around to ask people to get ground truth
... then use it on the 4 billion data set

hhalpin: where did you get the data?

<tinkster> nonfriends = acquaintances who would not describe themselves as friends (i.e. colleagues, business contacts, etc). Not necessarily enemies.

nathan: orginal study reality mining at MIT 5 year ago

hhalpin: how can I choose to grant to a company the acces to my location/calls/etc. and make sure the other can't?
... what kind of rich data can you get from mobile phones.

<hhalpin> so, the telcos get locations when calls are made.

<hhalpin> but what else is out there?

nathan: we didn't really use that much data from phone calls

<hajons> +q

nathan: used bluetooth scanning

<cperey> +1 on bluetooth

<tinkster> Bluetooth range is about 10m.

<cperey> category of mobile social network: proximity communities

nathan: to evaluate proximity of people in context (e.g. always clos at 3am sunday morning)

<hhalpin> so proximity data gathered via bluetooth is the most important.

nathan: other expermient with 10 000 bluetooth+GPS+ microphone on every ten seconds for some sec
... use the sampling to detect their context, the media they consume, etc.

<cperey> what is the name of this study?

nathan: over 85% of the random sample agreed to do it

hhalpin: needed with a special phone?

nathan: need the right kind of phone

<bblfish> I think newer versions of iPhone os will (already does?) allow background apps

hajons: one subject of the group is to define ontology to describe context

nathan: I work with raw sensor data
... not an expert on sharing context
... having a common vocabulary would make sense

<danbri> "behaviour signatures ... eg dinner with a person on a regular basis ... high likelyhood will report as a friend, ... true in africa, india, finland ... universal rules about our species"

nathan: there seems to be an overlap between the different experiments done around the world

<tinkster> bblfish, quite possibly. How else is one supposed to run an HTTP server on an iPhone? ;-)

<rreck> this is a great discussion

nathan: there could be some universal way to describe generic context e.g. having diner -> friend

hhalpin: you worked in very different places and found that there are some similarities.

<rreck> i think its time i turn my phone off when im not using it.

<hhalpin> So, it's in the PNAS paper this behavioral signature.

<rreck> thanks so much for presenting to us

<danbri> +1

+1

<Adam> +1

<karl> wonderful information

hhalpin: what is the best paper?

<hhalpin> 100 is PNAS paper

nathan: IEEE paper

<hhalpin> IEE paper is 10,000 mobile ambient experiment

<rreck> is IEEE available?

<rreck> yes please send the draft

<bblfish> yes please send draft too :-)

<karl> for the record, the draft must not be public

IEEE paper ? http://ieeexplore.ieee.org/search/wrapper.jsp?arnumber=4580547

<hajons> +1

<danbri> +1 for book draft

hhalpin: we are interested in helping researchers to share data. Can you share the data?

nathan: I have never seen "sharing agreement" work to my satisfaction.

<hhalpin> Sensor data from Noka: http://www.sensorplanet.org/

<hhalpin> Nokia, simple concerns - http://alpha.nokoscope.com/eb2/login

nathan: attempts never really materialized because of privacy implications

<rreck> we cant even determine how to share what my friends are to my satisfaction

nathan: the reality mining data is online and publicly available but is kind of an exception to the rule

<tinkster> http://reality.media.mit.edu/download.php - reality mining data

nathan: I like the idea of people having to cite my paper before thy use this data.
... I spent 3 years gathering these data.

hhalpin: we realized it is hard to get data and we hope improve it

<hhalpin> factions?

<hhalpin> companies?

<hhalpin> mobile datas?

nathan: many other groups have had the same agenda for many years.
... I can give you a list of these groups

<hhalpin> http://crawdad.cs.dartmouth.edu/

<rreck> im overwhelmed

<karl> head spinning

nathan: we saw differences in radius, phone frequencies and duration patterns, e.g. in Africa an average of 3sec.

<rreck> glad its not just me

<hhalpin> Not in near-term

<hhalpin> User-generated content in developing world!

nathan: we need to develop content for developing world

hhalpin: example?

<rreck> im sure that problem would be solved when the providers recognize it enhances the data they have

nathan: like a job posting liste, item to sell
... the current cost of it is to high right now

<hhalpin> http://txteagle.com/

<hhalpin> Mechanical Turk...

<hhalpin> Mobile Mechanical Turk for developing world.

<karl> Wonderful ! really!

<hajons> great! thanks

nathan: send me a reminder to send you the draft.

<rreck> good question harry

<danbri> "biggeset impact, trying to encourage legislation, and operators to change pricing schmes"

proposes ... ACTION: hhalpin to ask Nathan Eagle for the draft of his chapter

<scribe> ACTION: hhalpin to ask Nathan Eagle for the draft of his chapter [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action20]

<trackbot> Created ACTION-67 - Ask Nathan Eagle for the draft of his chapter [on Harry Halpin - due 2009-08-26].

<hhalpin> Nathan - data standardization is premature

<rreck> WOW, that was awesome

<bblfish> very interesting

<hhalpin> Meeting adjourned

<Adam> .aagg is me

<tinkster> FabGandon, aren't most people native speakers (of at least one language)?

hhalpin: we need to try to finish the Use Cases doc

rreck: we should focus on finalizing the use cases

hhalpin: may be less invited speakers ?

<bblfish> sounds good

<bblfish> can we also discuss the use cases on the mailing list?

<bblfish> so for example deal with each use case as a seperate mail, and work it out online?

hhalpin: we push over next month the use case document

<hhalpin> that would make sense...

<hhalpin> some of the use-cases have been discussed very heavily, and others haven't.

henry: take the use cases one by one and send email for each one?

<bblfish> yes

hhalpin: five at a time not to flood the list ?

<hhalpin> ACTION: bblfish to send some e-mail about use-cases out... [recorded in http://www.w3.org/2009/08/19-swxg-minutes.html#action21]

<trackbot> Created ACTION-68 - Send some e-mail about use-cases out... [on Henry Story - due 2009-08-26].

henry: if no one answers on a use case we just drop it

hhalpin: ok

rreck: the biggest issue raised by Nathan was privacy

<hhalpin> I like that metaphor that the doctor OWNS your health care information, just like the telco, and that information should be YOURS.

+1 fab ; heard Nathan say technology/standards were not the problem