Technological advances have transformed mobile phones into sophisticated gadgets capable to tell us when the next train arrives, to guide us when we lost and can even allow us to watch your favorite TV shows. Now, researchers have shown that this device can be used for something seemingly detached from SF: cell phone knows what we will do tomorrow! A team of computer scientists at the University of Birmingham managed to predict the future position of people just by looking at their mobile phones.
The margin of error predictions of scientists was only 20 meters, which has fueled concern that organizations advocating the protection of privacy. To predict where they will find the owners of mobile phones, scientists have devised an algorithm that predicts their future movements. The algorithm compares data analyzed data person close people, forming its social network.
Analyzing the places visited in the past and frequency of contact between friends, the algorithm can predict where the person will be in the future. For example, if two individuals who meet frequently visits a restaurant, it is very likely that the next time these two individuals while heading toward that area, they have targeted that restaurant.
The algorithm compares data analyzed data person close people, forming its social network. Analyzing the places visited in the past and frequency of contact between friends, the algorithm can predict where the person will be in the future. More information on http://www.gpsphone-tracker.net/.
The research was based on data generated by mobile phones to 200 people in Lausanne, Switzerland. Dr. Mirco Musolesi, study leader, said that the information extracted from the data generated by mobile phones is an interesting source of knowledge about human behavior. We proved that accurate predictions of the future location of a person can be enhanced when taken into account the moves it has made in the past, along with data generated by social group of individuals that make up that person.
Despite the fact that this system will benefit the marketing companies and police, organizations that advocate for privacy protection do not see with good eyes this technology. Researchers say the technology could be used by advertisers to create targeted advertisements, anticipating the places they intend to visit someone. Also, the technology could be used by companies to provide discounts personalized, real-time, depending on where it is estimated that it will find cell phone user. Also, scientists hope that this algorithm will be used in future for police to predict locations where crimes to be committed.
Efficiency of the system has been tested in several experiments Patwari and Wilson. They placed 28 so-called “nodes” of the network’s inside a hall, delineating an area of about 25 square meters, and then in a park with 34 nodes positioned along the perimeter of an area of 56 square meters ( devices were placed on plastic poles at the height of a man).
Since the presence of objects and people attenuate radio signals between nodes, Wilson and Patwari measured intensity radio waves from all nodes, so when areas were crossed by people moving, and when they were completely empty. They then developed a mathematical formula that allows a computer to convert these variations in intensity images of persons represented in the form of colored spots.