A mind-wandering programme capable of detecting involuntary lapses in concentration has been developed by scientists in the US.
The "ubiquitous phenomenon" of mind-wandering - or zoning out - is attributed to having negative effects on people's ability to learn and was estimated by scientists from the University of Notre Dame and the University of Memphis to occur 20% to 50% of the time with subjects in their tests.
In an effort to better understand this phenomenon and create an intervention method that could prevent it, the researchers created a way of detecting mind-wandering.
The study, titled Toward Fully Automated Person-Independent Detection of Mind Wandering, was presented at the User Modeling, Adaptation, and Personalisation (UMAP) conference in Aalborg, Denmark.
"Our lab (Emotive Computing Lab) has been studying ways to automatically detect and respond to students' cognitive and affective states like confusion, frustration, and boredom for almost a decade," Sidney D'Mello, assistant professor in the department of computer science at the University of Notre Dame and lead author of the study, told IBTimes UK. "In this research, we decided to focus on detecting mind wandering."
The mind-wandering detection system works by using an eye tracker to monitor a subject's eye gaze patterns, both when concentrating and when loss of concentration is detected.
Machine learning methods then use this data to establish which gaze patterns in individuals signify that they are mind-wandering.
Combining the eye tracker with the machine learning technique allowed researchers to predict mind wandering with an accuracy of 72%.
"Now that we can detect it with some accuracy, the next step is to trigger interventions aimed to dynamically restore attention when mind wandering occurs," D'Mello said.
"For example, the system can briefly highlight content missed due to mind wandering, trigger attentional reorienting messages, ask a quiz question and so on - there are many possible interventions that we are exploring."
D'Mello believes that the software could have major implications for high stakes tasks, such as those within military or aviation contexts.
Plans to develop this technology commercially are currently in the works, with discussions with the University of Notre Dame's Technology Transfer office ongoing.