Researchers from Murdoch University’s School of Engineering and Information Technology are working on e-learning software that can sense when students are frustrated or confused and adjust to their needs.
Named ‘Becky’, the program uses sensors which can be seamlessly integrated into a mouse or keyboard to pick up subtle physiological changes such as shifts in heart rate and skin conductance (when skin momentarily becomes a better conductor of electricity) to detect emotional states.
These emotional states, notably confusion or frustration, prompt Becky to offer the learner revision or additional guidance on the subject.
Her effectiveness was recently tested on groups of students working through a lesson on genetics, of which none had prior knowledge.
The students monitored by Becky showed a strongly significant statistical advantage over their counterparts and reported more mastery and confidence of the subject afterwards.
Dr Nik Thompson said the research was part of a larger on-going initiative to improve e-learning and make communication with computers more natural.
“A significant proportion of human communication is non-verbal, taking the form of vocal patterns, facial expressions or body language, yet computers are essentially blind to these cues,” he said.
“At present, e-learning technology is much less effective than having a human tutor, which can be traced to the missing emotional component. For example, we know that anxiety impedes concentration and confidence enhances motivation.
“An expert teacher doesn’t just provide facts. They recognise emotion, stress and confusion and spend half their time responding to these needs. So why shouldn’t a computer do the same?”
Dr Thompson said Becky worked like a standard e-learning program, only deviating when she sensed her learner was feeling anxious or uncertain.
He added that while other researchers in the world were exploring affective computing, their efforts often resulted in highly complex technologies. Murdoch, on the other hand, was committed to developing open, accessible and affordable technologies that could be used by non-experts.
“This is very much about practical computing. We want people to be able to take our tools and adapt them to existing e-learning programs to make them better,” he said.
Dr Thompson said his group had created a novel model to show how their technology could be integrated into off-the-shelf software, and noted that the e-learning sector generated $35.6 billion in 2011, and has been growing at a rate of roughly 10 per cent since 2006.