Thursday, May 30, 2013

Red dot theory

This might seem a bit off subject to start with, but bear with me, you might just get the creeps by the end.

Imagine if you bought some new kind of shoe.  A one off by some talented but anonymous designer that every time you stepped on the ground, it left a big red dot.  You go about your daily business and eventually someone else, someone who works out its a shoe doing this by the length of stride and dot patterns, decides to find out who you are, presumably to charge you for cleaning off all these red dots  everywhere.

The easiest way to do this is to follow the dots to a house rather than a shop, and see which house is used the most.  In the meantime you got wind of this strange person stalking you and decided to only use your red dot shoes when you went shopping.  You did this to retain your anonymity.

A scan of all the red dots tells your tracker some interesting things, that you mostly go into men's clothing shops and you stop and wait outside shops with guitars and phones and gadgets in the window.

Your tracker quickly builds up a picture of who you are, and the more red dot trails, the more they can reduce the numbers of people likely to be you down to one.  Perhaps you go to a football ground to play once a week.  By analysing when the dots were created, you can narrow the numbers of people in a city down from a million to maybe thirty.  By combining the location data such as your football match, with buying data such as where and when you bought some sports socks, it would not be that long before your tracker could stand outside the football pitch and point with precision to who you were.

All this data is PAST data, it's already happened, it's supposedly anonymous but still a trail of breadcrumbs that leads to you as you create ever more clues.  So what's this got to do with education?  Most students create vast breadcrumb trails every day, most put their real name and photo on their Facebook pages and their hobbies and interests are easy for others to see.  Many schools generate similar sophisticated tracking data they are giving away to their sponsor or host or website when they use a specific education package or free app online.  "Right students, write me a piece on this online word processor of what you did on your summer holidays".  It is most unlikely the data retains the students name on it as such, but its not difficult to marry up the different strands in the future and identify them.  All that data is worth something to somebody.  A very tiny something because there is so much of it.

At the same time your students are giving away information about themselves by using sites online, even ones specifically claiming to be morally correct with that data, that could later be used to reconstruct their movements and identify them by cross referencing with other supposedly anonymous data.  Again, it may not have their name on it but if they live in Huddersfield, read a lot of judo websites and were not online on some sites on the day of a big judo competition, it is possible to work out who they are.

As a school or college or even university, you would be prosecuted if you tried to keep such immense amounts of data linked to a students name, but because the companies do not specifically link the name, it is allowed, its only when the data is analysed, or mined, that the patterns come through.  That mining used to be very expensive, impossible even when people did it by hand but now computers can do it with ease, and at the point the mining becomes easy, the data used is not just that going forward in time, it includes the past, sometimes decades of it.

There is an old navy saying used much in the Eighteenth century as a toast when drinking together.  "To wives and lovers, may they never meet".  In this day and age the data , like wives and lovers, can combine and tell you much about the man...

So what can we do about this?  I haven't a clue...

I explained this to someone last year and they told me they didn't use the internet so were never tracked, they were anonymous.  I asked them how long they had owned a credit card or debit card for. 20 years.  Every item tracked dated and geo located...  When you add to this peoples opinions posted online, we and our students, are not anonymous any more.

As far as I know, there is not a global policy to delete all data gathered from students when they reach a certain age.  Its too valuable, and therefore not so anonymous as we think.  In 2012 an online company went bust, no money, no real buildings, computers or office furniture as presumably it was all rented, nothing but a file of data on its customers.  Almost purely on the value of that data, the company sold for over $100,000 .  How much will your students data be worth in the future?  Should we allow it to have a value in the economy at all?