Re: Privacy Jungle: Data Protection in Social Networks

On Jun 13 2009, Story Henry wrote:

>Very nice paper Sören. Being on this group is worth it just for  
>this. :-)

Thank you very much!

>You argue that most sites do not promote privacy in 4.2.4 .  At the  
>same time it seems that for many of the sites it really is an  
>important part of the way they function. People seem to be very  
>conscious that not everything they say is public on Facebook, and vice  
>versa that mostly everything is on Twitter. It changes the way people  
>behave on those networks. So there seems to be some tacit knowledge  
>that is not expressed verbally but that is well understood...

I'm unable to tell the users' awareness of how public or private the information they post online is. My own anecdotal experience is that users often act surprised when confronted with the amount of information that's available about them on a platform. I think the number of people changing their (very permissive) privacy default settings could be an indicator for awareness of privacy controls -- which would be very low then. Based on our data, we argue that privacy is not (yet?) a selling point. As such, more recently established sites score less well in terms of privacy and data protection controls.

>Twitter:
>
>twitter is  a micro blogging site. so it does not need privacy. Since  
>everything is quite clearly public, it should get full points on the  
>privacy socre. Since it is the fastest growing site that would argue  
>for your correlation of privacy and growth scores in 5.4

Indeed -- Twitter is quite different from other sites -- both by its functionality, its target, and its social structure.

>Initially you claim that you always use the same information on all  
>sites. In this section you argue that the site owners could play a  
>game of presenting more or less information to different users  
>depending on the sites perception of the type of user they are dealing  
>with (privacy fanatics, practical majority, exhibitionists)

Indeed, a site may present different information depending on the sensed privacy concerns of a user. However, this is not our cause here. We argue that those who actually bother looking up a site's privacy policy are those who are more concerned. And such user groups will find different information. Even if the site does not tailor its content -- the mere presentation of the privacy-related documents and notices hides them from the mainstream user (see the example of Friendster in Section 4.3 for instance).  Eventually, we need to see privacy negotiations to give users privacy choices by making available fine-grained controls for data release.

Sören

Received on Monday, 15 June 2009 08:48:30 UTC