American researchers say they´ve developed educational software capable of sensing and responding to the emotional state of its pupils.
The project was spearheaded by Sidney D´Mello, Assistant Professor of Psychology at the University of Notre Dame, who assembled a team of fellow researchers from the University of Memphis and MIT.
The group says that their pedagogical software, which they´ve dubbed AutoTutor and Affective AutoTutor, is able to assess not only what its students are learning but also how they feel about what they´re learning. Whether a student is enjoying the lesson or feeling frustrated, the program is designed to field incisive questions, analyze the learner´s responses and then determine how to move forward most effectively.
Mimicking the intuitive gifts of effective human teachers, the software is able to detect if the student is enthralled, perplexed or discouraged with his lessons by reading facial expressions and body posture. Using this data, the program can then dynamically adapt the content and strategy of the lesson to fit the user´s mood.
D´Mello explained that his group´s project was based largely on the idea of getting to the heart of how humans interact with computers.
“Most of the 20th-century systems required humans to communicate with computers through windows, icons, menus, and pointing devices,” he explained.
“But humans have always communicated with each other through speech and a host of non-verbal cues such as facial expressions, eye contact, posture, and gesture.”
“In addition to enhancing the content of the message, the new technology provides information regarding the cognitive states, motivation levels, and social dynamics of the students.”
As D´Mello points out, a large part of successful education derives from reading and responding to the psychological state of the learner and providing motivational strategies that engage not only logic and reasoning but also emotions.
And because emotional engagement is a two-way street between the teacher and the pupil, the program itself even shows emotion in response to the mood of the user, modulating the tone and facial expressions of an animated tutor to fit the emotional state of the learner.
Researchers have already tested AutoTutor on more than a thousand students. The results thus far, they say, have been impressive.
In experimental trials, AutoTutor was able to facilitate measurable improvements in learning equivalent to roughly one letter grade. This, say researchers, excels the didactic abilities of most new human teachers and rivals the abilities of even experienced professional tutors.
“Much like a gifted human tutor, AutoTutor and Affective AutoTutor attempt to keep the student balanced between the extremes of boredom and bewilderment by subtly modulating the pace, direction, and complexity of the learning task,” explained D´Mello.
Details this new technology will be published in special edition of ACM Transactions on Interactive Intelligent Systems.
On the Net:
- University of Notre Dame
- University of Memphis
- ACM Transactions on Interactive Intelligent Systems