I was watching a TED talk earlier: Jennifer Golbeck: Your social media “likes” expose more than you think . Although a few years old now, it raised some really interesting questions – looking at Facebook as an example, there are “huge amounts of personal data posted”, meaning models can be built, using the information provided, which predict the behaviours of the individuals posting that data.
But how many of us actually understand the techniques analysts use to pull out our information and predict our behaviours? I’m guessing not that many…
Well, Golbeck highlighted the importance of patterns of behaviour – small decisions which don’t say much individually, but, when you’re dealing with millions of people, build up to indicate patterns which can be used to predict everything from sexuality and intelligence to whether you are starting a family.
Also, be aware of who you’re friends with – homophily (I love that word!), aka the tendency to be friends with people who are like us can also be used to predict our character traits: friends with several ardent football fans? Chances are you might be interested in a season ticket. Are all of your friends posting pro-conservative literature online? Then it’s a fair bet you’ll be voting blue in the next election.
There is also a reason why stories are described as going ‘viral’. It seems that, through homophily, stories and behaviours spread in exactly the same way as viruses, with the same patterns seen in a viral outbreak, making mapping and extrapolating data even easier.
So, what can we do about getting control over this data (which is used for advertising and manipulation in ways we may not appreciate or benefit from)? There are legal possibilities, although it’s such a complicated area, and so amorphous, that useful control is very difficult to put in place (and there are also issues over censorship etc.). Could we put pressure on the owners of the social media platforms themselves? Well, they make a large degree of their profits from sharing this data, so their actual, enthusiastic involvement is unlikely (although, since this TED talk was broadcast, pressure has been placed on various platforms, particularly Facebook, to better police and control content on their sites, and it may be that this will be extended to content use).
Golbeck suggests using science – develop mechanisms which inform users what information their actions add to big data – so, for example, a message popping up to a user saying “you’ve just liked this or that post, we can now make a fair estimation that you like a drink on the weekend”. This would allow users to make informed choices about their actions, giving a direct, visible link between an individual action and this enormous thing called ‘big data’. If users wanted to keep their information private, they would have that option, albeit by choosing to not carry out a particular behaviour online.
Would this work? In theory, yes, I believe so. The technology is there to allow for this feedback to be given to users, even though the platforms themselves may not welcome it. I guess the bigger question for me is whether the people themselves would welcome it: many times when I’ve discussed these issues with friends I’ve been met with a shrug and a ‘well, Facebook/Twitter etc. is free to use so I have to give something in return. It just means I get advertisements”. And, to an extent, this is true. However, the lack of clarity about just how that data is used, and how far reaching the effects, not to mention the massive profits companies are making with no explicit consent from the user (with the best will in the world, who really reads the terms and conditions in depth?), means that, actually, greater (and clearer) communication to users about data usage would undoubtedly be a step towards true user choice.