Your Smartphone Can Tell If You’re Bored
A group of researchers looked at how people used their phones to figure out when they were bored, then suggested they go read a BuzzFeed article.
Smartphones can be mined for immense amounts of data about you and your habits.
Add “boredom detector” to the seemingly endless list of things your smartphone can do. A group of researchers say they’ve developed an algorithm that can suss this out by looking at your mobile activity, considering factors like the time since you last had a call or text, the time of day, and how intensely you’re using the phone.
The work, which was conducted by several researchers at Telefonica Research in Barcelona, Spain, will be presented at the UbiComp ubiquitous computing conference in Japan next week. The researchers found that looking at this kind of data gave a reliable prediction of boredom as often as 83 percent of the time. The researchers also went a step further by sending bored smartphone users an alert to check out an article on BuzzFeed—which people who were judged to be bored clicked on more often than people who weren’t.
While using machine learning to infer your state of mind is tricky, doing so reliably via your smartphone could be powerful. For instance, if an app were able to predict that you’re bored, and also knew where you were, it could try to feed you content it thinks you’d like in that particular context. Already at least one startup is trying to do something similar to this: Triggerhood, which built software that lets apps collect data about how the phone is being used, determines when is the best time to send you a notification (see “Smarter Smartphone Alerts Come in When You Want Them”).
For their study, the Telefonica researchers first determined characteristics of boredom by using an Android app to ask study participants to rate their level of boredom several times a day over two weeks. The responses were compared with other data snagged from the phones measuring things like how many apps they used, and how intensely the phone was used overall (both measures rose as people got more bored).
To validate the resulting algorithm, researchers built another Android app that concluded on its own whether the user was bored, and, when it did, sent an alert to their phone asking if they wanted to read an article on BuzzFeed’s news app. A separate set of study participants used it for two weeks, and researchers found that the people who’d been identified as bored were more likely to click on the alert to see the story, and to spend time looking at it, than those who were randomly sent an alert.
Tilman Dingler, a graduate student at the University of Stuttgart and coauthor of the paper, worked on the study as a visiting researcher at Telefonica last year. He says the researchers now want to figure out more about what kinds of content people might most want to see when they’re bored, and whether that might include learning activities, too, like improving your Spanish.
Getting people to agree to use an app or service that analyzes lots of data about their phone behavior could be tricky, though.
There’s also a question of how accurately the researchers can predict boredom, given that they collected their initial data by asking people repeatedly to report how bored they were. M. Ehsan Hoque, an assistant professor of computer science at the University of Rochester who codirects the school’s human-computer interaction lab, says that this may not capture “true boredom,” as our mental states are often subconscious. He says a more objective way to measure the same thing would be by just asking people repeatedly if they want to play a game on their phone, and noting how often they say yes and then how long they play it.
However, Hoque says he’s excited by the promise of the study as it indicates researchers are tapping into a mental state using the smartphone data.
“We know boredom leads to depression, so if you can infer the person is bored, you can do something about it,” he says.
Keep up with the latest in machine learning at EmTech Digital.
The Countdown has begun.
March 25-26, 2019
San Francisco, CA