Tuesday, Sep 27, 2022

Teaching model may soon be able to detect students’ emotions

The system will adapt the learning content based on student performance.

Ramkumar Rajendran,a research scholar at Monash Research Academy of IIT Bombay,is currently researching on the development of a mathematical model for “intelligent tutoring systems”,which could predict and address in real-time the emotions of students,such as frustration,while they interact with the system. The primary aim of the research,according to the academy,is to develop a model that predicts the ‘affective’ states of users like confusion,boredom,excitement and frustration,among others,and address them,so that the process of learning and teaching is more effective.

The system will adapt the learning content based on student performance,background and previous knowledge. Some of the teaching tools are designed to provide personalised learning content based on the student’s needs and preferences.

“The benefit of such a system is that students,who like to learn at their own pace with computer-based self-learning methods,will not drop out because of boredom or frustration,since they are constantly engaged by content adapted by the system,based on their affective state,” Rajendran said.

The model being developed by him is based on information available in the log file of ‘Mindspark’,a math intelligent tutoring system.

Subscriber Only Stories
UPSC Key-September 27, 2022: Why you should read ‘Mankading’ or ‘Rarest o...Premium
Why India has lashed out at the US over its F-16 package to PakistanPremium
Kurmi club: On national path, Nitish looks east, at UP’s Sonelal Pa...Premium
Interview: MD-CEO, Central bank of India | ‘During PCA years, none of our...Premium

The work has been tested only on students in the age group of 10-12 and cannot be generalised to all age groups. “We constructed a basic model and our research is in its nascent stages of real-time emotion identification,” he said.

According to experts,the methods that have been implemented so far in the intelligent tutoring system to predict the affective state include human observation,self-reported data of learners of their affective state,analysing data from physical sensors,face-based emotion recognition systems,mining the system’s log data and assessing the data from physiological sensors like galvanic skin response and ECG,among others.

Academicians,however,said with the exception of data-mining approaches,the other methods were at present not feasible in a large scale real-world scenario to predict affective states.

First published on: 14-11-2013 at 03:43:13 am
Next Story

Back in town,Chandigarh Book Fair gets off to a good start

Latest Comment
Post Comment
Read Comments