A Canada-based researcher has been looking into the science behind attraction, exploring whether a computer has the capability to learn a person’s preferences of Tinder.
Earlier this week, PCWorld reported that Harm de Vries, a post-doctoral researcher at the Université de Montréal, became curious about the possibility after hours of unrewarding Tinder swiping.
De Vries’ research method included plucking vast amounts of images from a number of popular dating sites.
The researcher then entered his preferences and monitored whether the computer was able to give correct determinations based on the information he provided.
He used existing images from profiles on Tinder and OkCupid throughout the experiment, taking over 500,000 in total.
After a disappointing first attempt, De Vries then tried the experiment with a larger number of images, the computer achieved 68 percent.
Check out the full PCWorld article here.