How online dating sites and apps hone their software based on user behavior—and misbehavior—to find your true romance
February 13, 2015 | |
You may have seen the where Tom Haverford makes 26 different online dating profiles to increase his odds of matching with every woman possible (after his nerd profile matched with his boss Leslie). You may also have watched someone swipe right on every single option until they run out of every candidate within 100 miles or make joke profiles just for a laugh.Preventing these types of misuse and play is a big job for online dating companies. Identifying problems and deciding how to fix them is crucial for users looking for love, but now it's good for business, too.In 2014 the online dating industry made . alone has 2.4 million paid subscribers. Even Tinder, heralded as more of a game than an actual dating service by many Millennials, will soon start charging for a to get a bigger piece of the online market. People once looked down on online dating, but now it is widely accepted and continues to grow in popularity as new mobile devices provide additional platforms. One in 10 Americans has used an online dating site or app, according to a 2013 from the Pew Research Center, and 59 percent think they're a good way to meet people.So how do these companies keep their products running to find you love? Mike Maxim, chief technology officer at , says the company is always making minor improvements to its algorithm to make the service better. "Most of the changes at this point are fairly small," he says. "The users have an expectation of how the site is going to work, so you can't make big changes all the time."The biggest problem they face on the tech side, Maxim says, is to make sure everyone can find somebody. In their algorithm that matches users with one another they use , which basically quantifies how much users have in common, along with their popularity and in-box messages. On any dating site, he says, a small subset of users will receive the majority of the messages. To even this out they look at the number of unread in-box messages and place users further down the match list if he/she has tons of them. The popularity metric (which isn't displayed on people's profiles) helps them match people with similar status on the site.Misbehaving users are a continuous battle, Maxim says, especially on a free site like OkCupid. To fight this, he says, they use computer and human defenses. Their software can detect if someone sets up multiple accounts, claims they are in a foreign country or exhibits bad behavior, and it can then flag their accounts for review or automatically disable them. OkCupid also relies on reports from its users to find misbehavior, Maxim says. Steve Carter, vice president of matching at , says they close 300 accounts per day that their "highly experienced, dedicated and slightly paranoid 'trust and safety' personnel" deem spam, also by using software and human intuition.Like many online dating services, OkCupid amasses large amounts of data on its users, which Maxim says it uses to improve its products and monitor if the site or algorithm needs fixing. OkCupid president and co-founder, Christian Rudder, publishes some of this data and insight on the site's blog, , admitting unabashedly that they experiment on users.The years of data collection have also made the sites’ matching and operating algorithms smarter. Carter says eHarmony recently added a machine scoring system that can automatically crop photos for different devices and tell users which images will be most successful with possible mates. This data can also help sites be more personalized, says Vatsal Bhardwaj, general manager of Match. Sites catering only to , , , and already exemplify this desire to find someone with a very specific type and tastes.Experts agree that mobile will define the future of the dating industry, but what effects that will have on information is unclear. "There are a lot of ways in which the sharing of information online may grow or shrink," Carter says, "that could fundamentally change the way people use the Internet to find a mate." You can bet that the learning algorithms will change with them.
Chomsky: We Are All – Fill in the Blank.
This entry passed through the Full-Text RSS service - if this is your content and you're reading it on someone else's site, please read the FAQ at http://ift.tt/jcXqJW.
No comments:
Post a Comment