There is growing evidence that information gleaned from online social networks can be processed and used in ways to gain an accurate profile of an individual.
The latest work involves a study of geographical clusters of photos that users upload to Flickr. Kazem Jahanbakhsh and colleagues at the University of Victoria tried to determine the hometown of Flickr users by only taking into account the geotags of photos they uploaded. People take most of their photos near their home, but estimating the location is difficult because they also are likely to take photos in clusters at other locations, such as holiday destinations.
The team developed an algorithm to separate the home location from the other clusters and found that it guesses reasonably well. "In 70 percent of the cases our algorithm has predicted the place of living of people with low error," Jahanbakhsh says.
Such algorithms will make it hard to keep track of privacy because there is almost certainly a market for these types of applications. However, abusive practices could lead to more stringent laws on using and processing personal information.
From Technology Review
View Full Article
Abstracts Copyright © 2012 Information Inc. , Bethesda, Maryland, USA
No entries found