Forget about what the Web knows about you. Those static facts - your social security number, address history, home phone, etc. - are so yesterday. The future is all about what you're doing right now - it's all about you in real-time.
Consider the smart phone. At any given moment, it - and therefore your carrier - knows within a few feet exactly where you're standing. It knows when you're stationary or walking - and the direction you're heading. It knows who you stood next to on the transit bus, that you walked through Washington Square today when a political rally took place. It knows what time you went to the polls to vote. It can detect your conversations and background noises. It knows where and when you took every photograph. And it watches your every text and e-mail message. For better or worse, much of this information is being monitored, crunched aggregated and analyzed.
Smart phones with integrated GPS, like the iPhone (which private investigator Steve Rambam likes to call "the little snitch in your pocket") are the next frontier for digital data mining, says Tom Mitchell, professor in Carnegie Mellon's Machine Learning Department. That's just one data source in a network that may include such things as your Web activity and even the EZ Pass RFID tag on your car windshield that pays the toll for you - and tracks your vehicle as it moves through toll booths, recording your direction, average speed and other data.
The application of machine learning to this data has great potential for social benefit, Mitchell argues in a recent opinion, Mining Our Reality, in the journal Science. For example, Google already aggregates anonymous geolocation data from smart phones to track traffic congestion along major roadways and to display those conditions on its maps. Similarly, Google Flu Trends analyzes real-time search results on influenza to predict where outbreaks are happening. Mitchell envisions a world in which that analysis will extend to your conversations and even facial expressions to increase understanding of human behavior.
To the lay person it's all just a bit creepy. Personal data could be used to help with human health - telling you that you came into contact with a person subsequently diagnosed with influenza - or to prevent you from buying health insurance. It could be used to change traffic lights to optimize traffic flows - or to track where you're going. It could be used to understand human behavior to improve it - or to manipulate it and exploit human weaknesses for marketing, polticial or other purposes.
The "risks to privacy from aggregating data are on a scale that humans have never before faced," Mitchell says. The problem is that protections and policies are lacking. While you or I might worry about personal privacy issues, the big loser from Mitchell's view is science - and society. He goes on to describe the many ways in which analysis of real-time personal data could have clear social benefits.
Alas, he says, until we "rewrite the rules of data collection, ownership and privacy," the potential for scientific advances from the use of this data won't be fully realized. He worries about a consumer backlash that shuts the door completely to scientific study of that data.
But everyday citizens may be more concerned about what overzealous marketers, political organizations or certain government agencies may be doing with it.
This story, "Your iPhone May Be Snitching on You" was originally published by Computerworld.