Keeping it Real – Reality Mining

April 2008 issue of Technology Review lists the top ten emerging technologies, and the list includes something called ‘Reality Mining’.

The founder of the, Sandy Pentland from MIT, describes reality mining in layman terms as: “is all about paying attention to patterns in life and using that information to help [with] things like setting privacy patterns, sharing things with people, notifying people – basically to help you live your life.”

The idea is to keep a database of profiles of people and track the locations of these peoples. This combination could allow for inference of patterns in social behavior. The cell phone has been the device used to gather the data. For example, in creating an accurate model of an individual’s social network, the MIT team monitored a person’s phone call logs and the nearness of this cell phone device to other peoples’ devices that has Bluetooth sensors (laptops, phones.) A statistical technique used in social sciences called, factor analysis, was used to identify patterns in the data and thus the team is able to create a social relationships map.

How useful is this? Back in April, the example given was to develop privacy settings by categorizing the person’s address book between friends, family members, acquaintances, or coworkers. The idea is to auto-create the privacy settings to allow certain individuals to view your personal schedule. An extension would be to add the location dimension. The phone would be able to predict if the person is within the vicinity of someone in their address book.

June 2008 and the Sunday issue of the New York Times features a company that has found an application that is based on Reality Mining. In the “Bright Ideas” section, an article titled “Predicting Where You’ll Go and What You’ll Like” shows a picture of the founding owners of Sense Networks, and below their image is a Blackberry with the map of downtown San Francisco. (Sandy Pentland was initially one of the founding co-founders of Sense Networks, and now gives advise on privacy issues.)

Meanwhile, the product Gregory Skibiski and Tony Jebara (both science majors) are providing a service to businesses and consumers that make recommendations based on the users’ locations. Rather using telephone devices, their models gathered data from taxicabs that were installed with GPSs. (By the way, the new model of Blackberry that will be released in July will be installed with GPS chips.) In developing models for Sense Network product, Macrosence, it required gathering very large sets of data that go back many years.

There are other companies now engaging in Reality Mining, including Microsoft. A spin-off company called Inrix uses traffic data to predict traffic patterns. Something similar to this is Path Intelligence which monitors traffic flow in shopping centers by tracking the cell phones in the area. That would be a useful application for those who hate to be stuck in holiday traffic jams.

Macrosense and Citysense are services that users of iPhones or Blackberrys can sign up for. The software is designed to help the user in decision making for places to go, or show them where people of like interests are going.

“What Will They Think of Next?”

- Analyn Revilla