Follow us on

Does your dating site know you better than you?

by Jackie Dent
12 March 2015

A research scientist specialising in the algorithms behind online dating sites asks some interesting ethical questions about the role technology plays when we look for love, writes Jackie Dent.
Serious online daters can spend weeks, months and even years online. They craft flattering stories about themselves, post attractive photos, trawl through the sites looking for someone they want to take out.  They send messages, they wait for messages, they get rejected, they get lucky and they go on dud dates.  Sometimes they fall in love.
But for all the effort that can go into online dating – and of course a good looking photo certainly helps – the reality is the outcome is more driven by algorithms than fate, says Dr Luiz Pizzato, who has spent the last six years specialising in people recommender systems that use patterns of online behaviour to work out who people like and who will like them back. 
Using similar technology similar to Amazon’s ‘Customers Who Bought This Item Also Bought’ section. If two people, say Bob and David, want to date Alice, then we assume Bob and David have similar preferences. We can then infer who Bob may like, based on David’s actions. However online dating has the extra complexity, which is whether Alice would actually like to date Bob or David.
While Pizzato, who worked closely with a major Australian online dating site and whose research is used by dating sites in China, says algorithms are good at comparing data to find people who might like each other; but there are ethical issues. Algorithms have the ability to impact on someone’s understanding of themselves.
“If the computer algorithms do not have any provision for serendipity, then it may simply enforce a wrong understanding of one's preferences. For instance, let’s say Bob has only looked at people of European descent.  This is a coincidence and not a deep view of Bob's true likes and dislikes but the algorithm may then assume a dislike for people from other descents,” he explains. 
“If Bob then blindly follows the algorithm recommendation and matching score, it might actually enforce a view that is not quite true for him. For instance, this would cause the algorithm to present a low matching score for Bob and everyone with a non- European descent. The problem this presents is that Bob may start defining his own preferences based on what the algorithm tells him.”  
On the upside, the sites also have the ability to make people rethink who they may like. For example, Pizzato says city-dwellers have strict preferences and tend not to look for people who live further than 10 km away.  But if they find their preferences for specific hair colour or height cannot be met, they will change who they are looking for.  “A lot of people don’t really know themselves,” he observes.  “You may say this is the kind of person I’d like to date.  You give all these characteristics and when you go the website, not only are you going to find different people, people that don’t match your descriptions, you might see there are fewer choices than you expected and you might have to broaden the preference, and try someone different.”
The scientist has never used online dating to look for partner. He met his wife when he was working on his PhD at Macquarie University.  Nonetheless, he still relates to the idea of having to find new relationships.  As a Brazilian who moved to Australia, he says it hasn’t been easy making new friends and in response, he is developing a mobile app which helps people find buddies.
When it comes to the commercial aspect of online dating, Pizzato says internet dating companies face the ethical dilemma of wanting to get couples to meet while still keeping them as customers.  “These are conflicting goals: if you know two people are meant for each other, would you suppress that information for profit? How long can you suppress that information until people give up the service?” he asks.  “The most valuable customer is the one that will be happy enough to use the system for a long enough time before they find their best match. In this way they leave the online dating service with a happy experience at the same time that provided enough revenue for the company.”


Follow The Ethics Centre on TwitterFacebookInstagram and LinkedIn.

Twitter-Logo.png Facebook-Logo.png instagram-logo-sketch-copy.png linkedin-logo.jpg