Researchers have takengeolocation apps one step further by developing technology that may predict where you are headed.
Where you've been and where you are may soon become passe, thanks to Mirco Musolesi, a computer scientist at the University of Birmingham who helmed the project.
He and his team created an algorithm that tracks your own mobility patterns and adjusts for irregularities by factoring the movements of mutual contacts in your smartphone.
Using 200 people who were willing to be followed, the algorithm proved to be quite accurate.
It predicted where a person would be 24 hours later by less than 65 feet. When the same system predicted a person's potential location using only past movements (without the movements of friends), the average margin of error was just under 3,281 feet.
Of the 200 people tracked, all lived within Lausanne, Switzerland.
Many were students and researchers, who Musolesi said are relatively predictable. However, he did say results of the study were insightful.
"We are essentially exploiting the synchronizsd rhythm of the city," Discovery News quoted Musolesi as telling Technology Review
No comments:
Post a Comment