Definitions - starting the policy blueprint

[Elias Bizannes]

Hi everyone,

I'd like to get some consensus on key terms, as I believe the way we
define terms like "data" and "portability" can fundamentally affect
what the entire concept means. I've looked through my university texts
of old such as when I did information systems (database theory),
statistics, and even marketing - and I am finding them different. Even
on the web, there is difficulty finding consensus on what these terms
mean.

So this is what I propose:

---------
Data is an object which is a representation of something in the form
of characters, numerals, images or sound which alone does not have any
meaning.

Information is when meaning is generated from data. It occurs when
creating context of data by associating it with other forms of data.
Formally put: Information is the association of different data, that
through the relationships between the data objects, generates some
form of meaning.

Knowledge is the processing of information to create insight.

Where with portability, the ability to use data in another system
independent of where the data is stored.
-------
Examples.

123382 is data

$123,382 is information (adding context to the data)

"$123,382 is what 90% of the population do not earn in society" is
knowledge, because you've applied the information.

Another example -

"joe" is data
"j...@email.com" is information
j...@email.com is the e-mail of joe schmoe (by associating the
information to an identity, you have created knowledge)

I think there are some profound implications on how we define these
terms. Once we can agree on the concepts, the next step is to group
real world things ie, is a photo data or information? is an e-mail
data? Then, we need to determine things like who owns data, who owns
information? This latter conversation I think we have found some way
to work that out (as we have talked about it in the previous thread).
But I am more concerned on how we apply concepts to definitions like
data. If we exclude photos and e-mails from the definition of data,
this can mean a whole lot of things that affects DP at its core

Thoughts?

http://wiki.dataportability.org/display/dpmain/Policy

[Steven Greenberg]

I propose that we promote "data portability" as an approach rather than a specific definition, help foster and promote the various technologies, and let the industry converge on the details over time.  We need to tell a story that is concrete enough that people generally understand what it means but allows for enough flexibility that companies can fit it into their business models.

I find it useful to think about the issue in terms of the value that each party contributes.  The user provides content that has some value.  The site creates new value by organizing that original content, transforming it, or combining it with stuff from others.  The user is entitled to some of that new value, but not all of it.

Thus, I propose that this group should have two rallying cries:

• The user should be able to take with them any content that they originally provided.  If I put it in, I should be able to get it back out.  This covers my address book, pictures, videos, information I typed into fields, blog posts, whatever.  There are a lot of ways for sites to satisfy this, but we need to get this meme out there so that users know what to ask for.
• The site should provide a way for the user to take with them some of the newly created value.  Here Be Tygers.  "Some" is the operative word.  How much, and what, are questions that each site will need to negotiate with their users.  Our goal here is to get the users to start asking.

Ok, we're not going to hear people chanting those in the streets (maybe in Mountain View), but I think that these ideas are concrete enough to be actionable but broad enough to allow companies to figure out the right answer for themselves and their users.  

I believe that a formalist approach will be a dead end.  Each site has a different mix of data vs. information vs. service, a different business model, and a different set of concerns.  

Just my rapidly devaluing $0.02.

    Regards,

    Steve

[Elias Bizannes]

"Just my rapidly devaluing $0.02."

Don't say that! Good perspective, I think that's a good call.

[Iain Henderson]

Yes, I think those definitions are sound and that this line of thought would be a helpful one to pursue.

Regards

Iain

[Dan Brickley]

On 11/03/2008, Steven Greenberg <gree...@puzzlingevidence.net> wrote:
>
> I propose that we promote "data portability" as an approach rather than a
> specific definition, help foster and promote the various technologies, and
> let the industry converge on the details over time. We need to tell a story
> that is concrete enough that people generally understand what it means but
> allows for enough flexibility that companies can fit it into their business
> models.

Well said. Very well said!

> I find it useful to think about the issue in terms of the value that each
> party contributes. The user provides content that has some value. The site
> creates new value by organizing that original content, transforming it, or
> combining it with stuff from others. The user is entitled to some of that
> new value, but not all of it.
>
> Thus, I propose that this group should have two rallying cries:
>
>
> The user should be able to take with them any content that they originally
> provided. If I put it in, I should be able to get it back out. This covers
> my address book, pictures, videos, information I typed into fields, blog
> posts, whatever. There are a lot of ways for sites to satisfy this, but we
> need to get this meme out there so that users know what to ask for.

Yes please. This is a good summary of why I and others have invested
time in FOAF and other SemWeb technologies; and in why we don't scope
those technologies as narrowly as "buddylist exchange".

A good illustration of the complexity tradeoffs here, is in thinking
about how a full representation of Flickr exports would look. Turns
out you can get just that here:
http://search.cpan.org/src/ASCOPE/Net-Flickr-RDF-2.1/README ... and
the number of descriptive vocabularies needed to capture that data is
substantial. I don't see it as DP's role to do the data definition,
... but to evangalise and champion sites that do let you get your data
back in a responsible way. I say "responsible" because "your" is an
oversimplification when it comes to social data; it often describes
several people.

> The site should provide a way for the user to take with them some of the
> newly created value. Here Be Tygers. "Some" is the operative word. How
> much, and what, are questions that each site will need to negotiate with
> their users. Our goal here is to get the users to start asking.
>
> Ok, we're not going to hear people chanting those in the streets (maybe in
> Mountain View), but I think that these ideas are concrete enough to be
> actionable but broad enough to allow companies to figure out the right
> answer for themselves and their users.
>
>
> I believe that a formalist approach will be a dead end. Each site has a
> different mix of data vs. information vs. service, a different business
> model, and a different set of concerns.

Absolutely. Couldn't agree more.

cheers,

Dan

--
http://danbri.org/

[Steven Greenberg]

Thanks for the kind words. If nobody objects, I'll post a slightly
expanded version of this on my blog so that we can hold the
conversation in public.

[Christian Scholz / Tao Takashi (SL)]

Hi there!

I just wanted to chime in with a blog entry I finally got to write today:

http://mrtopf.de/blog/web20/what-is-data-portability/

Maybe this is of help. Not so much for the policy but for the technical group I think it makes sense to maybe have different fields and scopes on which to work and in which different methods and standards might apply.

Probably it's good though to keep the discussion on the lists.. I will also post it to the steering group as it seems to be a broader topic.

Cheers,

Christian

--
Christian Scholz
Tao Takashi (Second Life name)
taota...@gmail.com
Blog/Podcast: http://mrtopf.de/blog
Planet: http://worldofsl.com

Company: http://comlounge.net
Tech Video Blog: http://comlounge.tv
IRC: MrTopf/Tao_T

[Phil Wolff]

We frame much of our talk in terms of departure, "taking" being the common verb. A tad negative.

Consider a site operator's point of view, for example:

When you install data portability, your enterprise:

• opens your site to accept new users and their onlives. 
  
• speeds the time from newbie sign up to productive and satisfying experiences
• describes your users more comprehensively and with fresh information so you can serve them better, even when they haven't visited your site
• incents your users to make your site more useful for their social networks
• simplifies your users' choices about privacy and disclosure
• creates new opportunities for partnerships and alliances with other DP supporters
• helps you create safer user experiences
• basks in the consumer trust and affection for the DP.org lovemarks
  

- Phil

p.s. Are tads a metric unit?

Phil Wolff
managing editor, Skype Journal
http://SkypeJournal.com
pwo...@skypejournal.com
skype:evanwolf
+1-510-444-8234 San Francisco
+1-646-461-6123 New York
+44 020 8816 8780 London
+852 8175 8107 Hong Kong
http://www.linkedin.com/in/philwolff
http://www.facebook.com/profile.php?id=724232370

[Steven Greenberg]

+1

A point I often make is that the walls around your data don't make it hard to leave -- they make it hard to come back.

Friendster being closed didn't prevent Tribes or MySpace.  MySpace being closed didn't prevent Facebook or Bebo or Hi5 or Tagged.  Facebook being closed won't prevent whoever is next.

    Steve

[Christian Scholz / Tao Takashi (SL)]

Good points indeed. And that's why I also put "import" and "sync" on the list of levels. If data is automatically synced it also means that there are less bare profiles.

-- Christian

[Mike Pearson]

Please refer to the Data Portability Definitions page before
suggesting any new definitions:
http://groups.google.com/group/dataportability-public/web/data-portability-definitions

These definitions have been used for several months now. Any new
definitions must be consistent with the existing definitions.

--------

Discussions should NOT occur on any individual's blog pages.
Spreading discussions over several personal blogs means that the
discussion is not transparent - there will be no single linear record
of the discussions and comments. That is what these groups / mailing
lists are for. They are publicly viewable, even if people have not
joined them.

[Phil Wolff]

This page
http://groups.google.com/group/dataportability-public/web/data-portability-definitions
has permanently moved to
http://wiki.dataportability.org/display/dpmain/DataPortability+Definitions

Thanks.

[Phil Wolff]

Mike Pearson wrote...

Discussions should NOT occur on any individual's blog pages. Spreading
discussions over several personal blogs means that the discussion is
not transparent - there will be no single linear record of the
discussions and comments. That is what these groups / mailing lists
are for. They are publicly viewable, even if people have not
joined them.

Me: I think it's great that you blog, and that you blog about dp/DP. I
do it sometimes [1] because I like the greater authoring flexibility
(embedding images, etc.) and want to share it with a readership that
doesn't live here.

I agree with Mike, though: It's vital we copy our work into a Google
Group or the wiki (http://wiki.dataportability.org) so our group mind
can borg it. When I get better bandwidth than I have today (at the
computer history museum in mountain view) I'll do that with my own
posts.

wiki tip: the confluent platform lets you create a new page as a blog
post.

- Phil

[Elias Bizannes]

"These definitions have been used for several months now. Any new
definitions must be consistent with the existing definitions. "

Those definitions were never ratified in the first place by any of the
action groups or the prior workgroup to my knowlesge - it was simply
added. So whilst I agree we should build on them because you have done
an excellent job compiling them in the first place, I think its
mistaken to think those current definitions represent things we have
already established. Now that the policy group is starting to come
alive, we need to formally review these definitions one by one and
come to a consensus that it is appropriate.

Re blogging: how about this as a solution: you are allowed to post on
your own blog but you must also post an exact copy into the wiki via
your personal space (you dataportability blog effectively). That way,
we can centralise discussions with people still getting the
recognition of the thought they have put in.

[Phil Wolff]

Hi, All. If you think a definition needs clarification, leave a comment , add a second definition for the term, or change it. This is a living document, not The Ten Commandments. You can't break a wiki. So have fun. And trust the community.

--

[Kaliya *]

On Thu, Mar 13, 2008 at 2:31 PM, Phil Wolff <pwo...@gmail.com> wrote:

Mike Pearson wrote...

Discussions should NOT occur on any individual's blog pages. Spreading
discussions over several personal blogs means that the discussion is
not transparent - there will be no single linear record of the
discussions and comments.  That is what these groups / mailing lists
are for.  They are publicly viewable, even if people have not
joined them.

Sigh it is really hard to hear this kind of strict dogmatic talk.  There is a range of ways people communicate and there are a range of ways technical communities choose to self organize. 

Blogs were innovated for a reason for people to speak their minds and express personal opinions.  They are an important forum for people to express themselves AND help grow and build community around technical subjects. They put the conversation 'out' into a wide web public.  They actually change the quality of conversation on mailing lists because people do less soap boxing on mailing lists and put out their detailed thoughts and opinions on their blogs.

This can actually increase awareness of the conversation - transparency to believe that EVERYONE who cares about these issues is on these lists and is going to take the time to read about them here is just naive.

The Planet Identity Blog has been an incredibly effective community discussion tool and has been key to the success of the identity community.  http://www.planetidentity.org

I hate to suggest yet another communication channels cause this community has to many of them already (hopefully there will be some major pruning). But perhaps an aggregate blog is a good thing for those at the core of this conversation.

I think there is a diversity of ways to communicate and a diversity of places to talk about these things.  Encouraging healthy patterns and inviting people to 'also' do something may be friendlier way of encouraging the sorts of behavior you would like to see.

I notice a lot on skype and also here that people who 'say the wrong thing' in the 'wrong space' are immediately told so in a way that sounds harsh and I think will make some shyer people afraid to speak up for fear of getting jumped on.

I hope that you can hear these words in the spirit they are being offered which is helpful advice from someone who has been working to build and grow a healthy community (as in network centric - non-hierarchical) open standards/big problem solving community.

Regards,
=Kaliya

 

Me: I think it's great that you blog, and that you b tolog about dp/DP. I

[Elias Bizannes]

I think Kaliya raises some good points

- blogging: creates greater exposure of the issues. As long as we can
translate the discussions, or encourage more people to contribute (ie,
make it required every person taht posts links it to a certain forum
here), then we actually benefit more. I also think this group is going
to rely on the intellectual capital of its contributors, and they need
to feel recognised for coming up with innovative new approaches.
Requiring them to not blog is a disincentive form participating,
especially when people reject their ideas here. Their blog is an
opportunity to voice their ideas without getting talked down

- re the being shy to speak up, that's a good point and I will
personally will be more concious of that as I am sure I may have done
that without the intention to discourage anyone but just to moderate
discussions in the rooms. We should probably also invest in more
documentation helping explain to people how we have set up things
here, so that when new people join, they know what to expect, and
where things should be.

[Julian Bond]

Elias Bizannes <elias.b...@gmail.com> Fri, 14 Mar 2008 21:28:12

>- blogging: creates greater exposure of the issues. As long as we can
>translate the discussions, or encourage more people to contribute (ie,
>make it required every person taht posts links it to a certain forum
>here), then we actually benefit more. I also think this group is going
>to rely on the intellectual capital of its contributors, and they need
>to feel recognised for coming up with innovative new approaches.
>Requiring them to not blog is a disincentive form participating,
>especially when people reject their ideas here. Their blog is an
>opportunity to voice their ideas without getting talked down

Good grief, people. What is all this talk about "requiring"? Or what,
what are you going to do? People will blog what they damn well please.
And quite right too as it's their blog, not yours.

DP Evangelism should be encouraging as much as PR as possible. 3rd party
people talking about DP on their blogs is a major part of this. DP
Evangelism should be encouraging this without any pre-conceived ideas
about what is an acceptable post.

Please let go a bit. There's too much control-freakery round here.

--
Julian Bond E&MSN: julian_bond at voidstar.com M: +44 (0)77 5907 2173
Webmaster: http://www.ecademy.com/ T: +44 (0)192 0412 433
Personal WebLog: http://www.voidstar.com/ skype:julian.bond?chat
No Wife, No Horse, No Moustache

[Mike Pearson]

I have no problem with blogging about DP, so long as people realise
they are personal.

> Blogs were innovated for a reason for people to speak their minds and
> express personal opinions. They are an important forum for people to
> express themselves AND help grow and build community around technical
> subjects.

But it's vital to copy/summarise any personal blog learnings into the
DataPortability.Action.Policy Google Group or the wiki (http://
wiki.dataportability.org).

Do we need a DataPortability.Action.Policy blog? Is there something
about the functionality, that we don't get from the other two tools?

~mike

[Dan Brickley]

On 15/03/2008, Julian Bond <julia...@voidstar.com> wrote:
>
> Elias Bizannes <elias.b...@gmail.com> Fri, 14 Mar 2008 21:28:12
>
> >- blogging: creates greater exposure of the issues. As long as we can
> >translate the discussions, or encourage more people to contribute (ie,
> >make it required every person taht posts links it to a certain forum
> >here), then we actually benefit more. I also think this group is going
> >to rely on the intellectual capital of its contributors, and they need
> >to feel recognised for coming up with innovative new approaches.
> >Requiring them to not blog is a disincentive form participating,
> >especially when people reject their ideas here. Their blog is an
> >opportunity to voice their ideas without getting talked down
>
>
> Good grief, people. What is all this talk about "requiring"? Or what,
> what are you going to do? People will blog what they damn well please.
> And quite right too as it's their blog, not yours.
>
> DP Evangelism should be encouraging as much as PR as possible. 3rd party
> people talking about DP on their blogs is a major part of this. DP
> Evangelism should be encouraging this without any pre-conceived ideas
> about what is an acceptable post.
>
> Please let go a bit. There's too much control-freakery round here.

Quite. If people feel the urge to centralise, how about taking a
half-day to install and test http://intertwingly.net/code/venus/ or
similar blog aggregator. Then the rest of us could offer RSS/Atom
feeds for dataportability.org-related discussions.

cheers

Dan

--
http://danbri.org/

[Elias Bizannes]

I think a feed is all we need to centralise a decentralised means of
posting. Yahoo pipes, or a feedburner account that pulls our blog
posts with a certain tag.

[Improbulus]

1. "Portability". Agree with the concepts of "import/export" and
"sync" being important aspects of portability.

2. "Data". Separately, on the concept of "data", I think the European
Commission Article 29 Working Group's definition of "personal data" is
a useful read - http://tinyurl.com/24hwns.

3. "Ownership". I agree with whoever said "control" is perhaps a more
accurate concept. In the physical world only one person can have their
hands on a physical object at one time, but in the digital world
obviously more than one person can replicate info/data once it's out
there, indefinitely - and it's controlling access and replication and
use of (relatively) easily-reproducible info/data which I believe is
the crux of it.

Imp.
http://www.consumingexperience.com/

On Mar 13, 9:27 am, "Christian Scholz / Tao Takashi (SL)"

[Elias Bizannes]

Why has discussion died on this very important issue?!!

Please read my latest thoughts: http://wiki.dataportability.org/x/JwkR

Concepts I am adding to the discussion (after considering the European
Commission's personal data paper; and Christian's excellent blog
post):
- we split "data" into identifiable information and multimedia.
> this has implications as identifiable information falls under the scope of identity, whereas multimedia falls under copyright and creative commons
- Defining data as 'free'
> Again, an important distinction. I think trying to determine data ownership is not only conceptually wrong, but too complicated for what we are trying to do

Whilst I agree with Greenberg's request we take a broad approach and
not try to over-define, I also think some of these concepts need
defining because they form fundamental premises to our entire
approach. As you see in my above link, I define data and information
not so much to make a point, but to build an argument which in turn
affects how we view DataPortability as a concept.

On Mar 18, 7:25 am, Improbulus <improbu...@gmail.com> wrote:
> 1. "Portability". Agree with the concepts of "import/export" and
> "sync" being important aspects of portability.
>
> 2. "Data". Separately, on the concept of "data", I think the European
> Commission Article 29 Working Group's definition of "personal data" is

> a useful read -http://tinyurl.com/24hwns.

>
> 3. "Ownership". I agree with whoever said "control" is perhaps a more
> accurate concept. In the physical world only one person can have their
> hands on a physical object at one time, but in the digital world
> obviously more than one person can replicate info/data once it's out
> there, indefinitely - and it's controlling access and replication and
> use of (relatively) easily-reproducible info/data which I believe is
> the crux of it.
>

> Imp.http://www.consumingexperience.com/