This page gathers notes for topical discussions in the Social Web XG. It may be rough and drafty. Feel free to help improve especially with links to related resources.
Widget Platforms and Addressbook / Contact APIs
W3C has a widgets framework targetting desktop and mobile installable widgets. It provides packaging and APIs for widgets written in ordinary Web formats (HTML, js, CSS etc). Meanwhile, efforts like OpenSocial (Google and partners, Apache Shindig etc) and Facebook have similar platforms but that operate through HTTP and websites. Both W3C Widgets and OpenSocial have significant adoption levels. The former largely in the Mobile space, the latter amongst a network of Social Networking sites initially assembled by Google, but also through Apache's opensource Shindig project.
In both OpenSocial and W3C Widget scenarios, application authors write .js code that deals with people, groups, properties of people, etc. In a mobile setting, there are addressbooks and discussion around widget apis for addressbooks. In the OpenSocial environment, there are REST APIs for addressbook and contact data, most recently merged with the Portable Contacts initiative.
This provides some background for a potential series of SWXG discussions: initially factfinding, but also investigation of schema and API sharing between these environments.
Who should be invited? (List-servs and People): @@
When should it be?: @@ June 17th, June 24th?'
(Note, Art Barstow of WebApps WG has requested that we delay this telecon till after their F2F June 9-12th)
What are the concrete goals?
Note there was concern that WebAPI's widget spec is getting close to finishing LC, and that it would be too late for suggestions for dependencies.
Connecting VCard, FOAF, Microformats, and Portable Contacts
In the last few years, a number of open standards have been produced for describing people and their personal contacts. In particular, for people, the primary contact format has been VCard, which has its own text-based data format. Two mappings (the VCard RDF note and the VCard Ontology) exist in W3C space, causing some confusion. Ideally these should be combined. But however, vCard is now a moving target, with a new IETF VCard working group is updating it and adding an XML syntax, and it would be useful to know how much vCard is changing. Another XML-based syntax is the growing work around [ortableContacts, which should be co-ordinated. FOAF, another popular portable vocabulary for social contacts, needs to be co-ordinated with VCard and PortableContacts, as well as microformat efforts around XFN. How can we get a simple yet extensible way of describing contacts that operates over both vCard and FOAF, and has both an XML and RDF syntax?
Who should be invited? (List-servs and People):
- Someone from VCard IETF
- Someone from Portable Contacts
- Brad Fitzpatrick (SocialGraph API)
- Renato Iannella (vCard RDF)
- TimBL (W3C)
- PeterMika (Yahoo!)
- Sandro (W3C)
- Ralph (W3C)
- Martin Hepp
- Dan Brickley (with FOAF hat on)
- Norm Walsh and Harry Halpin (vCard Ontology)
- XFN and hCard Microformat people (Brian Suda? Tantek Celik?
When should it be?: @@ May 27th, June 3rd, June 10th, June 17th?
What are the concrete goals?:
- A single RDF vocabulary (VCard RDF 2.0)
- Can we co-ordinate this RDF serialization with XML work in IETF VCard group?
- Should VCard RDF 2.0 merge with FOAF
- Hosting issues with FOAF
- How to make FOAF work with XFN? Homepage versus person issue again.
- Can we clarify relationship between Portable Contacts and vCard? Seems very compatible.
- Co-ordinate FOAF to handle both vCard and PortableConacts? XFN? FOAF Hosting issues?
- Would W3C blessing help any of these technologies?
Privacy and Context
Over the last few years we have seen an exponential rise in the creation and publishing of user generated content on the internet. The development of inexpensive devices, and the fact that Social Networking sites allow for rapid and effective transport of of personal information/media to a user's desired social network, being two key drivers.
Below is a non-exhaustive list of potential topics for discussion given the topic of 'privacy and context' in the social web. The following list of topics are proposed to help/inform the task force of issues pertaining to how contextual information can be used to enrich a user's experience of the web, whilst ensuring that a user is informed on how their information is used, so that their privacy may be protected at all times.
- Privacy, the amendment and removal of misleading information:
From a privacy point of view, Garlik, the online identity experts, recently undertook some research Garlik Privacy report that pointed out the failings in the UK's pubic sector when it came to adhering to Data Protection Act (DPA).
Could this incubator group undertake a similar task with popular social networking sites? It is understood that the DPA only applies to UK companies, but could a similar endeavour be undertaken looking at networking sites Terms and Conditions and Privacy Policies.
Should the task force aim to identify best practices when it comes to privacy policies in social networks? Should these best practices be tested out on existing social networking sites? How can we ensure that information about a user can always be removed, or corrected if it is somehow misleading?
Who should be invited? (List-servs and People):
Tom Ilube (CEO - Garik) ?? Someone (Facebook/other social networking site) ??
- Data ownership and unintended reuse.
Can an open social network protect the integrity of multimedia items by upholding copyright and licensing issues, as well as ensuring that media has not been doctored or reused without the owner's permission? Are there technologies which would help enforce the aforementioned values?
What techniques / technologies exist which would allow for licensing information to be embedded into multimedia? Perhaps we could get some input from the XMP community? Is this something which is of interest to this group?
Oshani has done some work on the detection of violations of the creative commons licenses in Flickr Images see Detecting Creative Commons License Violations with Flickr Images on the Web.
1. Unintended reuse of content and personal information
"Jennifer Ringley also achieved celebrity and indeed notoriety when she set up JenniCam, a webcam that recorded events in her living space, which made available on the Internet images of events in her daily life ranging from the mundane to the pornographic (cf. Jimroglou 1999). Initially she filtered out private moments, but for most of the experiment, which ran from 1996 to 2003 the images were unfiltered (http://www.arttech.ab.ca/pbrown/jenni/jenni.html)
If expectations of privacy have fallen, then people are more likely to be on their guard, -awareness is a sort of protection against surveillance. However, this does bring us to Bailey and Kerr's important point (2007) about the decline of reasonable expectations causing a related decline in legal privacy protection. They note that Jennifer Ringley originally wanted to present a realistic picture of the life of a young woman, to show that it was not like an episode of Friends. JenniCam certainly was neither as entertaining nor as false as that show, but Ringley could not keep control of her information, and soon pornographic images filtered out of the total set began appearing out of context. The total picture soon became the edited picture."
2. The doctoring of multimedia items to distort the truth :
Images of attacks on Beirut BBC news article
NHS trust faked MP visit picture BBC new article
The LivingKnowledge EU project will attempt to look at how doctored images/information left on the web can be used to influence opinion and bias. Would this be of interest to the group ?
Perhaps input from the DataCommons or the CreativeCommons community? The XMP folk, the embedded metadata community may have some things to say about this? Oshani...
3. Data ownership and Anonymization :
In a world of walled garden social networks, how can users of social networking site ensure that if a site ceases to exist, they will have all of their information
returned to them. Who owns the data you post to a social networking site?
Can we imagine a decentralised social network where people own their own data? How would this effect concepts of removing information from a social space, if one doesn't know who owns the domain which is hosting misleading information posted about them.
Should this working group be looking at Anonymization techniques used by social networking sites? I am not sure if they get a mention already, but they should definitely being giving input into this. There is lots of work on reverse anonymizing user data, e.g. netflix data has been de-anonymized. Jon Kleinberg mentions such problems in of his keynote presentations. Do such topics lead on to topics such as the one brought up by Tim Berners Lee in his No Snooping note?
- Context :
An investigation into the positive uses of context should help any discussion around how to protect users from abuse. An example domain where context based information has been reported useful is that of personal multimedia annotation.
Mor Naaman has done some work on used social networks to annotate events/multimedia items. We could bring in experts which have looked at Context based data, and how it can be used in a positive light. People in the UK's Memories for Life initiative could provide insight into how look at how context combined with content is being used to aid memory. This includes work making using of the SenseCam ( Hodges, S., Williams, L., Berry, E., Izadi, S., Srinivasan, J., Butler, A., Smyth, G., Kapur, N., and Wood, K. SenseCam: a Retrospective Memory Aid. Proc. Ubicomp 2006, pp. 177--193. )
Marc Davis has done work on how contextual information can be used to help identity faces/locations (Marc Davis, Michael Smith, Fred Stentiford, Adetokunbo Bambidele, John Canny, Nathan Good, Simon King, and Rajkumar Janakiraman. Using context and similarity for face and location identification. In Proceedings of the IS&T/SPIE 18th Annual Symposium on Electronic Imaging Science and Technology Internet Imaging VII. IS&T/SPIE Press, 2006. )
Given that context is not always a scary thing, what are the sources of context which can be gathered from the Web outside that of personally curated information. We know of information on our desktop computers: Nepomuk (Semantic Desktop) work, as well as location based info from devices and the W3C Geolocation API, and information about our social life, and this could be combined with weather information, new stories, etc. Would this working group be interested in looking at what information can be found on the Web? Perhaps this could lead to the detailing of what information could be turned into Linked Data on the Web.
- Technologies :
This task force be looking at what technologies would be need to create a privacy aware social network? What would be nice would be detail what a privacy aware decentralised social network would look like, and what technologies could be used to realise such an endeavour.
FOAF+SSL (Henry Story) PushBack technologies (Michael Hausenblas) FOAF (Danbri) XMPP, etc