Anonymised data isn't nearly anonymous enough – here’s how we fix it

It’s too easy to identify people in supposedly anonymous personal datasets, but there are ways we can protect our privacy, says Yves-Alexandre de Montjoye