In BriefResearchers have developed an algorithm that can detect with 84 percent accuracy when children, instead of their parents, are playing with a cell phone.
Next Level Parenting
Shiny screens have a way of appeasing wailing children far more effectively than the old jingling keys standby. However, 21st century parents have to weigh the benefits of near-instant placidity with the very real possibility that their toddler could unknowingly max out the AmEx buying gummy bears on Amazon. Thankfully, new software developed by researchers from the University of South Carolina and China’s Zhejiang University could help make render this particular parenting dilemma moot.
The researchers developed an algorithm that measures a user’s interaction with the mobile device and can reliably tell if the user is an adult or a child. If the software detects a child, it can automatically block applications like retailers or email platforms, as well as inappropriate websites.
In order to construct the algorithm, the team developed an app that tracked users’ finger movements — recording metrics like the surface area of a tap, pressure applied by a finger, and length of swipes. The researchers gathered data from a group of children ages 3 to 11 and a group of adults between the ages of 22 and 60 as they unlocked the screen and played a numbers-based game on the phone.
Their new age-detection software proved to be 84 percent accurate in determining whether a user was an adult or a child with just a single swipe. That accuracy shot up to 97 percent after just eight swipes.
The algorithm hasn’t been integrated into an operating system yet, but the researchers will present their technology at HotMobile, a mobile tech conference, where it could gain some traction with developers.
While cybersecurity is an ever-present concern in the age of mobile devices, many make the mistake of only considering external threats. The reality is, an inquisitive three-year-old could be almost as damaging as the latest data hack.