Using Biometric Sensors for Detecting User Emotions in Educational Games

By Heather Maclaren

Affective Computing is a relatively young field that is concerned with modelling, expressing, and reasoning with emotions. At UBC we've focused on the first of these tasks, and are attempting to model a person's emotions during human-computer interaction. Our testbed (which many in LCI have had a hand in at some point) is the Prime Climb educational game. We are aiming to produce an affective user model that combines the evidence of both the causes of a person's emotions and and their observable effects into a single emotional assessment.

A previous talk presented the overall structure of the emotional model and focused on how the goals that the student wanted to achieve would affect the emotions that they experience. This talk will look at our work towards the part of the model that reasons about the student's emotions using evidence collected from biometric sensors. In this case the sensors are used to measure facial expression (eyebrow movement), galvanic skin response, blood volume pulse and respiration.


Back to the LCI Forum page