They’re watching you: Non-members of social networks are still products

A few people I know have removed their Facebook profiles, in the interest of gaining back some sense of privacy. In fact, I even know a couple of people who have never signed up to a social network in the first place. Sometimes I envy these people, just because they are taking a stand against a growing phenomenon that is rapidly turning personal data into a commercial product. Still, I have often wondered how much of a difference it makes if you choose not to belong to something as ubiquitous as Facebook or Google+. Now, a group of boffins at the University of Heidelberg in Germany, have released a paper analysing how much information can be gathered about non-members in a social network.

Social networks like Facebook and Google+ are gathering data about members and their relationships all the time. If you are a Facebook user, like myself, and you have avoided entering as much personal information as you can, Facebook simply monitors your relationships with your friends and their interactions with you and can actually calculate things about you without you providing a shred of information. Currently, it is clear that things like gender, political preferences and the types of relationships that you have can all be predicted with an alarming degree of accuracy, based on the information that your friends provide about themselves.

That goes with the territory: sign up and ‘all your base are belong to us’. Social networks also collect information about us from outside of the networks themselves. Every site you visit that has a Facebook like button or a bit of Google code embedded causes your system to send information back to these networks. That means that right now, Facebook and Google both know that you are visiting Memeburn, and they’re using this information to build a profile about you. But, if you’re one of my very rare friends, you may feel that you are still relatively safe from this little privacy invasion, since you have purposefully avoided signing up to these social networks. In vain hope, you assume that these social networks are possibly unable to work out who you are.

Here’s where this latest study becomes concerning. Up until now, most studies like this have been based around what social networks can work out about their members. This study is about what can be discovered about non-members, for which a social network has built a ‘shadow profile’. Based on information provided by members within a social network like Facebook, and the probability with which a member may make data such as their address book available to the social network, the team were able to calculate that they could calculate relationships between non-members to an accuracy of around 40%.

Since the study was only carried out within Facebook’s environment, where the availability of contact data is much more restricted, it is likely that companies like Google (where many of its users are taking advantage of Gmail), these calculations are likely to be much more accurate. The analysts have pointed out that they only relied on a very limited set of data being available. However, if you take into account the other information that social networks gather such as age, gender, education, location and work place, the profiles they can build on non-members is pretty impressive.

Katharina Zweig, one of the analysts who worked on the study has commented, “Overall our project illustrates that we as a society have to come to an understanding about the extent to which relational data about persons who did not provide their consent may be used”. This insightful statement illustrates the underlying problem with social networks and the way in which they mine big-data. It is understandable that information is gathered about users who have knowingly agreed to the terms and conditions of a service. However, we are currently in a position where information is gathered about people who have not consented to, and are often actively opposed to, this kind of data mining.

I have seen numerous commentators on the Net, including some on Memeburn, who tout the argument that targeted advertising is an improvement and that privacy advocates are blind to the advantages of identity profiling. Often neglected are the implications for security and identity theft, political censorship and control and the power that we are handing to corporates to control the information that we have access to. I believe that studies like this show that we need strong legislation against this kind of activity, and there needs to be global co-operation in enforcing the protection of identity information.