Researchers developing boredom-sensing algorithm for smartphones to read human mind

An algorithm developed by a group of researchers in Spain will soon enable smartphones to sense if their users are getting bored. It will be able to do so by tracking the users’ mobile activity and taking into account factors like time since last call or text message, the time of day and how intensely they used their devices.

Researchers at Telefonica Research in Barcelona and Germany’s University of Stuttgart found that they could give a reliable prediction of smartphone users’ boredom 83 percent of the time by looking at this data.

A set of 54 volunteers was asked to log details like frequency of cell phone usage together, some apps they used, etc. together with their levels of boredom over a two week period. Thus, data gathered established a direct correlation between boredom and phone usage.

It was seen that boredom causes smartphone users to spend more time texting, calling people or going through their apps.

Research also says, “Being bored makes mobile phone users more open to consuming suggested content.”

Researchers then sent push notifications of Buzzfeed articles to the phones of these volunteers when the algorithm predicted they were bored. They were seen to respond to such notifications 20 percent of the time and read an article for at least 30 seconds in 15 percent cases.

The response rate to such notifications was significantly lower if sent to volunteers when they were not, according to the algorithm, bored. The links were clicked open the only 8% of the time with a reading rate of 30 seconds or more at 4%.

In a blog post, authors wrote, “While we certainly don’t feel that recommending Buzzfeed articles will cure peoples’ boredom, at least not for the majority of them, the study provides evidence that the prediction works.”

The team of researchers now wants to know more about the nature of content most people want to see when they are bored, said Tilman Dingler, a graduate student at the University of Stuttgart and coauthor of the research.