Story at a glance
- In an experiment, study participants were shown several images of human faces while researchers collected data on their brain activity.
- Researchers fed brain activity data into a machine learning model, which then was able to predict which new faces would be attractive.
While technology companies work harder to try to determine our preferences using complex algorithms, scientists are studying brain activity data to predict our preferences, as well as to shift to “mindful computing” where computers and neuroscience can teach us about ourselves. A team from the University of Copenhagen has been working on that problem exactly and recently presented a paper at The Web Conference 2021 showing some promising results.
Study participants were shown a large number of human faces and told to make note of which ones they found attractive. The electrodes attached to the heads of the participants simultaneously recorded their brain activity. The researchers then could compare the brain activity among the participants and train a machine learning model to try to predict which new faces a person would find attractive.
Turns out, the model was able to do just that.
“Through comparing the brain activity of others, we’ve now also found it possible to predict faces each participant would find appealing prior to seeing them. In this way, we can make reliable recommendations for users - just as streaming services suggest new films or series based on the history of the users,” said senior author Tuukka Ruotsalo of the University of Copenhagen's Department of Computer Science in a press release.
For researchers, this technique might be helpful in situations where it is difficult to ask participants what their true preferences are. What someone may say is their preference may not match up with their behavior.
“Due to social norms or other factors, users may not reveal their actual preferences through their behaviour online. Therefore, explicit behaviour may be biased,” said Michiel Spapé, one of the authors on the study. “The brain signals we investigated were picked up very early after viewing, so they are more related to immediate impressions than carefully considered behaviour.”
But don’t expect to be able to use this kind of technology any time soon because it's challenging to develop. Besides the danger of misuse of personal data, there could be ethical issues with interpreting this kind of data for decision making.
This data wouldn’t be easy to collect and analyze and may not be scalable to be useful for advertising and marketing. But it could be important in helping someone understand themselves better, with many more potential uses. This is what the researchers call “mindful computing” and what they see as the potential for greater self awareness.
“The electrical activity in our brains is an alternative and rather untapped source of information. In the longer term, the method can probably be used to provide much more nuanced information about people's preferences than is possible today,” Ruotsalo said. “This could be to decode the underlying reasons for a person's liking of certain songs -- which could be related to the emotions that they evoke.”
READ MORE STORIES FROM CHANGING AMERICA