Illustrated By Austin Watts.
Much like gluten-free beer or tank tops with built-in bras, online dating is one of those things that's better in theory than in actual practice. Sites and apps promise to use the massive amounts of data on users' tastes and preferences to find them the love of their life. But, as anyone who has ever been brave enough to hand over their love life to a series of remote servers knows all too well, things can get pretty unsavory, pretty quick. The Internet, apparently, is filled with ego-centric a-holes (who aren't even that cute!) — and, no matter how many success stories you hear from your friends and those eHarmony commercials, it's all too easy to lose hope in the bleak, depressing realm of online matchmaking.
But, maybe we shouldn't delete our Match.com accounts quite yet. Professor Kang Zhao, a researcher at the University of Iowa, has come up with a new algorithm that promises to improve the capacity of websites to provide users with desirable (and compatible) matches. Whereas most of the existing systems are based solely on what users say they find attractive, Zhao's "hybrid" algorithm combines data about both your tastes and your attractiveness to others. That is, information about who you're attracted to, who you are messaging (and who messages back), and who finds you attractive, is compared to that of other users, creating a scoring system that places each user on a multi-dimensional grid of compatibility. In other words, the new system could go a long way toward connecting users with potential matches that they not only find attractive, but who are likely to message them back. Take that, OKCupid.
(Business Insider)