Researchers out of Spain have developed a machine-learning model that they say can recognize when mobile phone users are bored, then push content at those users.

“Being bored makes mobile phone users more open to consume suggested content,” the researchers, three of whom are from Telefónica Research in Spain, wrote. (They’re associated with Spanish telecommunications giant Telefónica; the paper’s fourth author is from the hciLab at the University of Stuttgart.) Their paper will be released at a ubiquitous computing conference in Japan next week (h/t MIT Technology Review).

The researchers explain:

If mobile phones are able to detect when their users are killing time, i.e. when attention is not scarce, then they could suggest a better use of those idle moments by:

— recommending content, services, or activities that may help to overcome the boredom;

— suggesting to turn their attention to more useful activities, such as revisiting read later lists, going over to-do lists, or participating in a research survey; or

— helping the user to make positive use of the boredom, such as using it for introspection, since mental downtime is essential to reflection, learning, and fostering creativity

Introspection and mental downtime, ha! Instead, the researchers’ app pushed BuzzFeed articles at users the app determined were bored. “We chose the BuzzFeed app as suggested content, because (1) the app caches articles, so that the study did not rely on permanent availability of an internet connection, and (2) its content is designed to be interesting to a broad audience.”

Turns out that “participants were significantly more likely to open and engage with suggested content on their mobile phones when our algorithms predicted them to be bored…These findings are significant, as they show that automatically detected boredom may be an ideal way to deal with people’s increasingly scarce attention.”

The details on how, exactly, the researchers’ algorithm assessed boredom are interesting and covered in full in the paper, but in case you’re wondering what, exactly, people who are bored and on their phones do:

Our participants tended to be more bored the more time had passed since receiving phone calls, SMS, or notifications, and the less time had passed since making phone calls and sending SMS. However, the volume of notifications received in the last 5 minutes is likely to be higher when being bored.

Being bored is also correlated with more phone use: the screen was less likely to be covered (which, for example, happens when the phone is stowed away), more apps were used, the last unlocking and checking for new notifications happened more recently, and the volume of data uploaded was higher when our participants were bored. Interestingly, the amount of data download and battery drain were lower when people were bored.

Related to demographics, male participants tended to be more bored than females, and boredom was higher for participants in their 20s and 40s and lower in their 30s.

Boredom was more likely the later it was in the day and the darker the ambient lighting conditions.

Finally, apps that most strongly correlated with being bored were Instagram, email, settings, the built-in browser, and apps in the ‘other’ category. Apps that correlated most strongly with not being bored were communication apps, Facebook, SMS, and Google Chrome.

So add “boredom detection” to the list of strategies that publishers will one day be able to use when they’re deciding how to send out push notifications. And this sort of contextual detection could be important for wearables, which are both a rich potential source of behavioral data and a notification machine on your wrist.

The full paper is here, the authors’ blog post explaining it is here, and the Technology Review writeup is here.