Online dating is hotter than ever. According to the Toronto Sun, visitors and revenue more than doubled between 2007 and 2012, increasing from $900 million to $1.9 billion. The BBC says that there are more than 5 million online dating site users in the United Kingdom. If you’re one of them, you’re probably used to filling out your online profiles much the way you would a job application.
Only Part of the Story
You don’t do it to intentionally mislead anyone. You’re just trying to accentuate the positive and eliminate the negative. Jennifer Gibbs, an assistant professor of communications at Rutgers, tells ABC News that people want to create Internet dating profiles that will stand out in the vast online dating market. But doing so means the algorithms most dating sites use won’t be able to accurately match people up. Omitting information, for example, means your dating site can’t recommend matches that share your actual interests. For instance, if you neglect to list on your profile that you enjoy online gaming and consistently download games from sites such as iwin.com, your dating site won’t connect you with other online game enthusiasts.
Or maybe you’re portraying the “you” you think you should be—or wish you were—instead of the one you are. An example would be if a woman who consistently dates and contacts blonde men of average height were to list in her profile that she was attracted to tall, dark men, a contradiction that’s obvious to anyone. It’s discrepancies like this that have caused researchers to look to companies such as Amazon for better ways to make a match.
A Calculated Match
A team based at the University of Iowa, led by management sciences assistant professor Kang Zhao, has developed a dating site algorithm that works more like the one Netflix uses to recommend your next movie selection. After completing a case study of online dating, Zhao’s team applied the data to create an algorithm that they believe will improve the response return rate for online daters by connecting people who more accurately match up with each other.
How would something like that work, you ask? In a nutshell, instead of giving any credence to the candy-coated wish list you filled your profile with, Zhao’s algorithm looks at the types of people you’ve contacted to determine your taste, and rates attractiveness based on how many responded and didn’t respond. Essentially, the new algorithm understands that actions speak louder than words. In the case of the woman who says she goes for tall, dark men, the old algorithm would continue to recommend what she asked for: tall and dark, with the responses continuing to be hit and miss. But the new algorithm would ignore her stated preferences once it recognizes that she actually prefers blondes under 6 feet tall.
If you’re uncomfortable with an algorithm having a mind of its own where your love life is concerned, you don’t have to pull out of your dating site just yet. Although Zhao and his team have been contacted by dating services that want to put the new algorithm into action, it hasn’t happened yet. On the bright side, you can always follow the algorithm’s lead. Emulate its penchant for ignoring and disregard the fact that a computer program knows you better than you know yourself.