The Pop-Up Ads of the Future Will Be Tailored to Your Mood

In the not too distant future, computer and gaming consoles will read your emotions, tailoring how they display information to better suit your mood. It will be able to do this because of how you type, or, rather, how differently you type when you feel different ways.

Researchers at the Islamic University of Technology, led by A.F.M. Nazmul Haque Nahin of the university's Systems and Software Lab, have created an algorithm that correlates the rhythms and patterns of typing with mood. Their findings are published in the journal Behavior & Information Technology.

The purpose of this program is to make computers "more efficient and friendly," the researchers wrote. "Computer systems that can detect user emotion can do a lot better than the present systems in gaming, online teaching, text processing, video and image processing, user authentication and so many other areas where user emotional state is critical."

Video games could adapt to gamer's emotions by changing aspects like the music selection or graphics, while teaching programs might change their style and interface, all depending on keystrokes. This, the authors argue, would inherently make the user's experience better.

The specific keyboard dynamics that researchers looked at included timing, like dwell time (how long the finger pressed on the keys), flight time (the timing between each letter you press while you type), and how many characters a user presses in 5 seconds. How you write (like, how many back spaces were pressed or words were deleted), also gives away a user's emotions.

The study was done in two parts. First, participants had to retype two paragraphs from Lewis Carroll's Alice's Adventures in Wonderland, and then pick which emotion they felt while typing. The second part of the study collected text samples from users as they went about doing their normal business on the computer, while software ran in the background and collected data from the keyboard.

Every 30 minutes, the participants were asked to pick their emotional state. The moods tested were joy, fear, anger, sadness, disgust, shame, or guilt. The researchers also added neutral, none of the above, and tired, just in case the participants didn't feel any of the aforementioned seven emotions.

The computer program was able to correct identify a person's mood roughly 70 percent of the time. The emotion it was able to identify the most was joyful (87 percent), the only positive option mind you, while sadness was the least identifiable. Sadness was only correctly guessed 60 percent, or 71 percent of the time, depending which part of the study.

The researchers admit that additional studies will have to be done because of the variables encountered during data collection. First, their sample size was small: only 25 people were tested between the ages of 15 to 40. Second, participants were less likely to input data when they were in negative moods. Lastly, efficiency with computers is another variable that would explain slow typing that doesn't necessarily mean the typist is sad.

If there are issues getting funding for another study, the researchers might consider contacting the advertising industry. Tailoring ads to moods seems like a lucrative possibility—or inevitability.