IGFWiki

Contents 1 Draft Project Description 2 What needs to be done. (big phases) 3 Prior and related Work 4 Meetup in Berlin?

Draft Project Description

Creating modular human readable Privacy conditions on websites

Instead of the long legal text of the "Terms of Services" for services (like facebook, gmail, hotmail, etc.) which everybody just clicks "Yes i agree" but nobody really understands the conditions, we want to develop a way to standardize these texts and then design a set of icons and create a "human readable" version for each aspect.

Once implemented this will allow to (a) to push for better standards, (b) raise awarness amongst users thereby (c) make Rights a relevant positioning / branding point.

Two design aspects seem particularly important in this context: The interpretations of the Rights to the context and needs of specific internet services needs to be organised in a way that (a) the standards can evolve (stakehodlers can initiate updates and amendments, as well as new standards for new applications) and (b) they should be modular (so the service provider can choose between several standardized Rights aspects).

What needs to be done. (big phases)

1) Decide whether our standarization use a) only text b) only icons c) both text and icons

2) In case we decide to use icons, to agree on and choose one icon modelization to work with.

3) Get several "terms of services" from different websites, licenses and so on. Try to make it as diverse as possible, from different companies, different products, different degrees on privacy restrictions, and so on.

4) From the study material obtained in 3) generate some analysis, identifying: a) paragraphs that are more difficult to understand b) good paragraphs which don't seem to need adaptation and why c) common privacy violations that can be found in those "terms of services" d) very specific privacy violations related to specific companies/services.

5) Synthesise those analysis in several lists which state things like a) legal terms that need to be rephrased b) legal expressions which need to be reconverted into a more understandable language. c) long sentences which can be safely reduced to a more simple expression d) key privacy concepts that need to survive the standarization and be understood by the general public e) legal technicalities that don't need to survive the standarization problem

6) Test these first results with our different targets: general public, Company representatives, web designers, graphic designers (if we are using icons), lawyers, policy makers, lobbyists.

7) Refine our synthesis based on the feedback obtained in 6)

8) wrap all the information into a modular system so it can be more easily used by third parties.

9) Test this modular system with our targets

10) Spread the word.

Prior and related Work

0. a great concrete example of human readable TOS: aviary.com/terms

left side legal text (lawyer-readable) -- right side human readable

1. A first proposal on iconography for privacy policies by Mary Rundle on the IGF in Athens.

2. Also already in 2004 the European Article 29 Group - where the European Data protection Commissioners meet and make recommendation - has clarified, that they would consider layered policies a plausible approach. Here is the Annex with an example - worthwhile for a quick view.

3. Matthias Mehldau, a young law student and ccc-activist from Berlin has developed a set of privacy-icons under a cc-license (article is in German, but icons and legend are English). There is also an English article about this.

4. OECD endorsed their privacy statement generator, which does not (yet) support layered policies, but features modularization.

similar generators are available (not tested):

• www.website-privacy-policy.com/
• www.dmaresponsibility.org/PPG/
• www.professionalprivacypolicy.com/

5. There is some research effort into this as part of the PrimeLife-Project. PrimeLife does do (amongst others) research in the field of HCI (Human-Computer-Interfaces) for privacy and also has some (limited) funds for implementation. It is the follow-up of PRIME, where you may already find a couple of deliverables looking into this.

6. Isotype Model, an approach based on the concept of Isotype, by philosopher Otto Neurath and artist/designer Gerd Arntz. The idea is to make a series of basic icons which can be articulated in a sort of visual language. So you have an icon for "man", and an icon for "factory". If you want to make an icon for "factory worker" you draw the icon of "man" with the icon for factory inside the drawing of the man. Here you have a link to a book by Neurath which describes the system. And here you have some examples of isotype icons that Arntz draw.

There is a more modern rendering of the model, developed by Rosa Llop and David Casacuberta, in which they used the isotype model in order to present a history of how techno music developed in different cities.

The isotype model can give us some methodology in order to find the best icons to work with.

7. PICOL, the attempt to develop a Pictorial Communication Language for electronic communication, started by Melih Bilgil and others in Germany.

8. The Platform for Privacy Preferences Project (P3P), an early attempt to standardize machine-readable privacy meta-data. Submitted to the W3C, but more or less dead by now. This was in fact the original inspiration for the Creative Commons licenses, which now serve as our inspiration. Some follow-up in the W3C on interoperability is done by the Policy Languages Interest Group (PLING).

9. A best practice in terms of rating system based on human readable icons is the PEGI (Pan-European Game Information) www.pegi.info/en/ . This is a great example on how an Industry can come out with easy to read and standard information which can really help consumer to make informed decision

Meetup in Berlin?

One opportunity for a meetup could possibly be in Berlin in the beginning of April, where we are going to have our next Privacy Open Space alongside with the re:publica web2.0 conference.

Discuss | What Links Here | This page was last modified on 2 July 2010, at 07:57. |