A prediction algorithm that uses location tracking, communication patterns, app usage, and other metrics was developed by U.K. researchers to track and predict usersâ€™ movement with up to 24 hours in advance.
The 200 volunteers in the 18-month experiment proved the prediction algorithm was less than 20 meters off in determining where users would be within the next 24 hours. Even friendsâ€™ patterns were used in the process because tracking their whereabouts could influence you into joining up with them.
Winning a Nokia-sponsored Mobile Data Challenge this year with their algorithm, scientists at the University of Birmingham believe they can further refine the data prediction process for better recommendations on ads and marketing campaigns that could be rolled out by companies.
Knowing that location tracking is something that most mobile operating systems and devices currently support, the issue of such data being used by companies is still controversial. The prediction algorithm was tested on “students, researchers, and people that are fairly predictable anyway“, says Mirco Musolesi, a computer scientist at the University of Birmingham who led the study.