The American Association for the Advancement of Science met in Boston last week to survey and discuss, erm, the advancement of science. Lots of interesting stuff going on with the boys and girls in white coats. One particular piece caught my eye. It concerned the position of mobile phones although I guess, in principle, it might extend to any device which we carry around with us that regularly connects to a network.
Song Chaoming of Northeastern University was able to acquire the phone records of 50,000 mobile users. He then developed an algorithm which, when applied to the data the records yielded, was able to predict with, on average, a 93% degree of accuracy where exactly that individual was likely to be at any given moment. In no case did the algorithm produce a result that was less than 80% accurate. Sam Spade is going to need to find a new way of earning a living.
Now at one level this degree of certainty about people’s movements is not especially surprising. Our lives tend to be filled with regular patterns of movement. But here we see what a single (and one imagines) well-intentioned social scientist was able to expose about us as individuals without him having to leave the comfort of his own computer terminal. I’m guessing all he had to help him were a few PhD students and that they did not have access to all the tools and tricks that “Big Data” have at their disposal.
Here’s the thing: not that long ago we had a fierce debate in the UK about the desirability, or otherwise, of having a national ID card scheme. I have no brief to defend the way the last government wanted to roll out their particular proposals, but its major opponents managed to conjure up a terrifying picture of an overbearing state which knew everything about everything we did, sometimes before we did it. I’d say we are already there, or very close, except it’s not just the state which has or is acquiring that sort of capacity.
If you add together what can be derived from mobile phones, WiFi log ins, GPS data, credit and debit cards, loyalty card schemes and, in London, our Oyster cards, congestion charge or other transport payments, not to mention CCTV, we are pretty much tied up and tracked from the cradle to the grave. And while I wouldn’t want to overstate the case I do think there is something unusually important about location data. It unlocks and points to so many possibilities, not all of them benign. If it is location data pertaining to kids things can get much scarier.
In the same field of research I came across the work of another chap who was also from Northeastern. Step forward Alessandro Vespignani. He’s an epidemiologist who has created a programme called GLEAM – Global Epidemic and Mobility Model. Vespignani has chopped up the world into little pieces and, using various bits of publicly available data about school holidays, people’s movements, some of it apparently obtained from social networking sites, he was able to predict where on the planet a particular strain of flu bug would start to do its dastardly work. In most cases he was right to within a week and he was never out by more than a fortnight.
This second case study is a marvelous illustration of how Big Data can work in ways which are likely to produce a public good, but both remind us of how increasingly small and confined our lives are becoming thanks to the internet and the many devices which hang off it in one way or another.
In the event of an outbreak of a deadly virus, I imagine if GLEAM was able to hook up with the mobile phone databases around the globe the World Health Organization might translate itself into an agency that had to authorise the sale of every airline ticket, from anywhere to anywhere. In some respects we’re not so very far off that already at least in relation to flights to and from the USA where people without any criminal convictions are being refused permission to land in or fly over US territory because of something in their past which links them to types of behaviour or organizations the US authorities don’t like.
I’m starting to feel a little claustrophobic. I’m also cross at the way various interest groups were able to derail initiatives emanating from bodies such as the UK government while appearing to be indifferent towards, silent about or unable to do anything to halt similar practices by companies quoted on the Stock Exchange.
In other words private enterprise is getting a free pass while public enterprise is getting locked up. In relation to my point about public agencies, I concede that my argument is not helped by the fact that, for example, the police rarely seem able to explain with any precision exactly why they want this or that type of information to be stored indefinitely and always to be made available to them instantly, just for asking. They think if they mention the “t” word everyone will or ought willingly to surrender everything without even troubling to ask why. Anyone who knows anything about the News of the World scandal look away now.
It is not easy to see a way forward for those who can see the creative and positive uses of Big Data, but we need to try to work one out. Strong and independent regulation has to be part of the answer. The draft data privacy directive currently wending its way through the European institutions is too flimsy a vehicle in the absence of a big, powerful transnational data protection agency capable of standing up to the might of the internet giants.
Desiderata 