Modeling Student Affect from Causes and Effects
Palestrante(s) : Cristina Conati (Computer Science, University of British Columbia)In this talk, I will present a computational model of user affect to recognize and adaptively respond to multiple user emotions during interaction with an educational computer game. Our model deals with the high level of uncertainty involved in recognizing a variety of possibly overlapping user emotions by probabilistically combining information on both the causes and effects of emotional reactions. I will start by describing the performance and limitations of the model when using only causal information. Next, I will illustrate the addition of diagnostic information on user affective valence detected via an EMG sensor, and present an evaluation of the resulting model.
monique.baron (at) nulllip6.fr