Students develop Cornell-specific dating algorithm just in time for Valentine’s Day

You see that adorable boy every time you grab your morning coffee at Libe Café and you wonder: What does he do on Friday nights? Could he be vegan? Maybe he’s a Republican? Might he be my … perfect match?

A group of Cornell students is hoping to help students find their special someone just in time for Valentine’s Day using an algorithm they’ve created and survey data from students about their Cornell-specific likes and dislikes, as well as personality characteristics. More than 3,400 students have already filled out the questionnaire and will receive an email with their perfect matches on Feb. 12.

“We wanted to do something that would be a service to our fellow students,” said Jeff Liu ’20, a government and computer science major in the College of Arts & Sciences and a member of Cornell Business Analytics, which organized the survey. “We thought maybe we would get 500 people signed up.”

Some Cornell questions include:

• What is your favorite place to eat on campus?
• What do you do on Friday nights?
• What’s your favorite first date spot at Cornell?

“If you and a potential match have a favorite café, that’s probably a good sign, whereas if you love CTB but your potential match hates bagels, then that’s probably not that great,” Liu said, adding that the other survey questions are more typical personality/psychology questions.

From left, Victor Odouard ’21, Jeffrey Liu ‘19, Samantha Taylor ’22 and Jamal Hashim ’22 created an algorithm to help Cornell students find their “perfect match.”

“As Cornell students, we’re so busy that by the time we get our work done and spend time with our friends, we don’t have much time to look for a relationship,” said Shashank Vura ’19, who filled out the survey. “And it’s hard to ascertain at your first meeting, whether this is someone you will get along with or have things in common with.”

Before creating the algorithm, team members did lots of background research and also surveyed 100 happy couples to find what questions seemed significant and what combinations of traits resulted in more “perfect matches.” Being too similar to a potential mate is not always a good thing, Liu said.

Victor Odouard ’21, a College Scholar major in the College of Arts & Sciences studying math and social sciences, said team members worked hard to refine their questions, come up with a method for scoring pairs based on compatibility and then deciding how to make pairings based on these scores.

“We wrote code to match people automatically and then have an automated email sender, so we never see people’s names or net ids,” Liu said. The deadline for filling out the survey is Feb. 11 and people will receive an email Feb. 12 with the name and net id of their perfect match, or maybe the top two or three.

“One match for everyone is a little limiting, and sometimes your No. 1 match doesn’t work out,” Liu said. “For instance, one of my friends said she did something similar in high school and got matched with her brother, so we’re trying to avoid that.”

It wasn’t hard to encourage students to sign up, team members said. Much of their marketing was done by word of mouth or on social media.

Samantha Taylor ’22, an engineering major, helped design the questions and then create posts for Facebook and Instagram. “In just over a week we were able to spread the word to thousands of students and create a buzz,” she said.

After the emails are sent Feb. 12, the Perfect Match folks are pretty much out of the picture, although they have planned a Valentine’s Day bowling event at Helen Newman for anyone who took part and wants a non-threatening way to meet with their match.

“I imagine most people aren’t going to show up, but it’s come to the point where we can’t possibly host everyone,” Liu said.

“I’m absolutely floored (at the response), completely in awe,” Odouard said. “I’m trying to put together everything that we did in my mind so I can try to replicate this in future projects.”

A month or so down the road, the student team wants to reach out to the pairs to find out how well the matching worked and see if there are tweaks they can make to the algorithm for next year.

For Vura, who says the highlight of his Valentine’s Day is sometimes the box of dark chocolates he gets from his mom, he’s hopeful he will at least meet a new friend or someone he wouldn’t have met without the help of Perfect Match. And who knows? Maybe it will be more?

Cornell Business Analytics formed in 2017 and is made up of students with majors ranging from engineering to computer science to business. The organization helps companies and non-profits tackle projects requiring data analysis and has worked with Lime Bike, a major car maker and yogurt company, as well as smaller organizations.

Students, to sign up for the survey, visit http://love.cornellbusinessanalytics.com/