The University of Toronto’s Student Newspaper Since 1880

Share on facebook
Share on twitter
Share on email

How a matchmaking algorithm paired up thousands of hopeless U of T romantics

Behind the Aphrodite Project
Share on facebook
Share on twitter
Share on email
KRISTAL MENGUC/THE VARSITY
KRISTAL MENGUC/THE VARSITY

If you are tired of Tinder, Bumble, Hinge, or the general flakiness of online dating, meet the Aphrodite Project: a student-run initiative that aimed to find the perfect match for students through a matchmaking algorithm. The Aphrodite Project is a free online matchmaking service that worked to bring two users in contact based on their responses to a questionnaire, with openings in a new Pandemic Edition until July 17.

Over 3,000 University of Toronto students participated in the Aphrodite Project in February, hoping to find their perfect match before Valentine’s Day. One of these students was Kathy Liu, a second-year undergraduate at Rotman School of Management studying business management. 

“It’s weird sometimes when I’m talking to [my match],” Liu said to The Varsity about her experience with the project. She said that he told her that he feels “like [he’s] talking to [himself] because [he] would respond exactly the way that [Liu] would respond.”

But what were the Aphrodite Project’s origins, and how could it help users find ‘the one’?

Who launched the project, and how was it funded?

The brains behind the Aphrodite Project at U of T are Aiden Low and Denise Yeo. Both students were in their third year at the time of the project’s execution, and both were on exchange from the National University of Singapore (NUS).

Low, a computer science student, has been on exchange at the University of Waterloo, while Yeo, a computer engineering student, has been on exchange at U of T. According to Low in an interview with The Varsity, their goal was “to be able to connect students and to build a better community.”

“Very rarely do we get a chance to meet people outside of circumstances that we are forced into, like classes or extracurricular activities,” Low said. 

The project was funded by the Innovation and Entrepreneurship Practicum Award, granted by NUS. The students received $9,000 to start the project, according to Low. In 2019, Low and Dana Lo, a psychology student at NUS, completed the first trial of the Aphrodite Project in Singapore.

Low considered the first trial a success and decided to adapt the questionnaire to students in Canada, running the project at the University of Waterloo and U of T.

How does the project match applicants?

The Aphrodite Project is based on the Gale-Shapley algorithm, which is associated with the 2012 Nobel Prize in Economics. It is a matchmaking algorithm that matches people based on their personality. 

The participants chose between a platonic or a romantic pool of pairings. Then, they answered a questionnaire with over 60 questions regarding their sexuality, cultural values, and personal preferences. Participants reflected upon questions such as their degree of engagement in politics and even how often they thank the bus driver before getting off the bus.

“We look at your deal breakers,” Low said. The program narrows down a pool of potential matches for each participant based on these deal breakers. Then, it compares the similarities between participants to generate a compatibility score. To determine this score, the program accounts for each participant’s answers to the questions as well as their typed list of hobbies and interests from the end of the questionnaire. 

According to the data report released on April 30, 3,269 U of T community members participated in total. The vast majority, 2,906 participants, were open to a romantic match while only 363 searched for a “purely platonic” relationship.

Most participants — 2,476 — were in their first to third years of study. An additional 430 fourth-year undergraduates participated, along with 113 students in their fifth year of undergraduate studies or beyond, 145 Master’s students, 30 PhD students, and 75 alumni.

A slight majority of the participants — 1,625 — identified as cisgender women, and 1,611 identified as cisgender men. Twenty-five participants identified as nonbinary persons, five identified as transgender men, and three identified as transgender women.

Thomas Siddall, a nonbinary person who is a fourth-year undergraduate with a double major in international relations and contemporary Asian studies, expressed to The Varsity that they were discouraged to see the low representation of LBGTQ+ participants upon seeing the data.

According to the report, the majority of participants — 2,827 — identified as heterosexual. 247 identified as bisexual, 114 identified as homosexual, and 45 identified as pansexual. Twenty-one identified as asexual, and 15 chose the option of “other.”

“We often like to think that queer… people are highly visible,” Siddall wrote, “but when you see metrics like this it can be disheartening.” 

They wrote that this may stem from a “fear of backlash” for LGBTQ+ people to participate in events like these, even with the anonymity of participants. The data report, semi-private for U of T participants and students, also used anonymized data to present trends in participants’ openness to cannabis use, “no strings attached sex,” experience with childhood or adolescent emotional trauma, and other attitudes.

However, the program did not share the participants’ responses to these questions to their matches. It only shared the participants’ name, email, hobbies, and interests.

In order to protect the privacy of participants, the organizers released a document titled “Data Principles and Protection Policy” that asserts that the “anonymized data will never be sold to any entities.”

What were some participant experiences?

Liu was satisfied with the results of the algorithm and considers her match to be compatible with her interests and preferences. 

“I would definitely try [the Aphrodite Project] again next year,” she said to The Varsity. “It’s a fun questionnaire… If your match isn’t successful, then you didn’t lose anything, and if your match is successful, then great, you just met someone that you can connect with.”

As feedback, Liu thinks that the two pool options of platonic or romantic do not take ambiguity into account. She believes that some people might be open to the possibility of a romantic relationship but may not be actively pursuing one. 

However, not everybody had a positive experience with the Aphrodite Project. Dina Dong, a fourth-year student double majoring in international relations and cinema studies, was frustrated by the idea of matching based on the key preference words at the end of the questionnaire. She thinks that just because people have the same preferences and hobbies, it does not mean that they view their passions similarly.

“My hobby is photography,” Dong said. “But the way I see photography is very different from how other people might approach [it].” 

Dong talked about the differences between pursuing photography as a career and pursuing it for aesthetic purposes only.

“It’s the whole human element that I think the algorithm can’t really capture,” she added. “Personally, I love the idea of organic growth between people, and I think it’s something that I very much would like to cherish, and I don’t want to lose it to a questionnaire.” 

A psychology professor comments on the project

The Varsity spoke with Geoff MacDonald, a professor at U of T’s Department of Psychology, to examine the promise of the project from an expert perspective.

MacDonald said he read the announcement of the project on Reddit shortly after Yeo and Low posted it on February 9. 

“I actually made a couple of comments on there because I was curious what they were up to,” MacDonald said.

He questioned whether matchmaking questionnaires similar to the Aphrodite Project’s, along with other dating services like eHarmony, are more effective than random chance. In one of his comments, he linked to a research paper co-authored by Samantha Joel, an assistant professor at Western University and former PhD student in his lab.

The research paper, which analyzed data from participants taking questionnaires before participating in a speed-dating event, concluded that “compatibility elements of human mating are challenging to predict before two people meet.”

Even while using “cutting-edge statistical methods and a vast catalog of psychological variables that have been widely cited in the field of relationship science,” the co-authors reported that questionnaires could not reliably predict romantic desire between two people.

However, though the Aphrodite Project may not guarantee romantic success, it did provide opportunities for many people to make connections, noted Low in an email to The Varsity.

Low’s response to MacDonald

In direct response to MacDonald, Low highlighted that the research paper itself noted that “the science of romantic relationships has much to learn from other prediction sciences before scholars can fully address this vexing and timeless question [of romantic desire].”

“Similar to the conclusion they reach,” wrote Low, “we don’t (and probably cannot) propose that we understand everything behind romantic desire – love is messy. Are our efforts worth it? Through the pandemic and so much happening this term, we’ve had friends and students getting back [to us] about their adventures in dating – sometimes fun, sometimes painful.”

Low also highlighted the positive effects of running the project, saying that he has heard from students who “are toughing out the pandemic” with their matches as well as from those who have “found meaning from a chance for reflection or in the data report.”

“In the end, being able to make that bit of positive impact for friends and students keeps us going,” he continued. “Love is messy, but it’s also trying, reaching, sometimes failing, and trying again. And as Joel et al. implies with their conclusion, let’s keep trying, and reaching, and learning – to keep exploring this vexing, timeless question and hopefully contribute in building a more loving world.”

The future of the Aphrodite Project at U of T

If you didn’t find your match this year, there is no need to worry: Low and Yeo’s pandemic edition is collecting questionnaire responses until July 17, with plans to announce results on July 19. 

“Right now we are still thinking of how to make [the Aphrodite Project] sustainable because we can’t sell data,” Low said in the interview. “That’s in the privacy policy. So we’re racking our brains on that.”

Low and Yeo are willing to hear feedback from the participants on their experiences, so they can continue to improve the algorithm. 

“People’s feedback would really help [us]… optimize for chemistry and make things work,” he continued. 

When asked if they would like to send a message to the participants, Low said, laughing, “I guess we could just say that if [the matches] work, we are happy to share our addresses for wedding invites.” 

Disclosure: Dina Dong is The Varsity’s Volume 141 Video Editor and was the Volume 140 Photo Editor. Adam A. Lam served as The Varsity’s Volume 140 Science Editor, and while he and co-author Ana Pereira matched together via the Aphrodite Project, no romantic couples were formed in the writing of this article.