Fast company logo
|
advertisement

MIND AND MACHINE

This AI matchmaker could mine Twitter to find you a date

Researchers built an algorithm that could spot a love match from public social media data.

This AI matchmaker could mine Twitter to find you a date

[Photo: Jenna Jacobs/Unsplash]

BY Jackie Snow1 minute read

A new deep learning algorithm called CoupleNet analyzes the tweets of two users and estimates how compatible the duo would be. It’s a different approach than dating sites and apps like Match.com or Tinder that rely on their data and algorithms to predict whether two users will “like” or text each other and then facilitate communication between possible lovebirds. CoupleNet, on the other hand, looks at public social media accounts for both explicit and implicit clues that two people could be romantically suitable.

Overview of COUPLENET model architecture illustrating the computation of similarity score for two users. [Image: courtesy of Yi Tay]
To train the algorithm, the team behind CoupleNet created a dataset of almost 2 million tweets pulled from over 4,600 likely couples. The “couples” were found by identifying public accounts who sent messages like “good night baby love you :-*” to each other. After training, CoupleNet could identify real couples in the testing data better than other recommendation tools that were given the same data. “To the best of our knowledge, our approach is the first data-driven deep learning approach for our novel relationship recommendation problem,” the researchers write.

CoupleNet could also explain itself. The algorithm was able to spot what mutual interests between users it thought suggested a good match, like tweets about K-pop bands. But it’s more than just word matching. CoupleNet could find pairs correctly even when there are no obvious matching signals. The researchers write that “we believe ‘hidden’ (latent) patterns (such as emotions and personality) of the users are being learned and modeled in order to make recommendations.”

advertisement

This type of tool could one day be turned into a “Who to follow”-type feature, but for romance. The paper also points out that their method likely overcomes issues that crop up on dating sites, like deceptive self-presentation, harassment, and bots. Meanwhile, tech companies like Facebook and Tinder continue to develop on their own AI-powered matchmaking efforts–and have incredible troves of data to mine. But unlike those carefully guarded datasets, the researchers behind CoupleNet offer a method that anyone can use on publicly available data. They’re also planning on releasing their own data set, so any machine learning-inclined cupids out there can build their own matchmaker tool.

Recognize your company's culture of innovation by applying to this year's Best Workplaces for Innovators Awards before the final deadline, April 5.

PluggedIn Newsletter logo
Sign up for our weekly tech digest.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Privacy Policy

ABOUT THE AUTHOR

Jackie Snow is a multimedia journalist published in or on National Geographic, Vanity Fair, The Atlantic, Quartz, New York Observer and more. More


Explore Topics