Use of personality profile in predicting academic emotion based on brainwaves signals and mouse behavior

• Main Authors: Azcarraga, Judith Jumig, Ibañez, John Francis I., Jr., Lim, Ianne Robert, Lumanas, Nestor B.
• Format: text
• Published: Animo Repository 2011
• Subjects: Personality and emotions; Brain—Magnetic fields; Computer Sciences
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1274
• Institution: De La Salle University

The academic emotion of learners is difficult to predict using EEG data, unless these brainwaves data undergo some extensive pre-processing operations. However, we show some evidence that it can be predicted somewhat more accurately for certain personality profiles. Twenty-five (25) college students were asked to use a math tutoring system while their brainwaves signals and mouse-click activities were being captured. Brainwaves signals were recorded using an Emotiv EEG device while the mouse behavior was based on the number of clicks, the duration of each click and the distance traveled by the mouse. The personality of the learners was evaluated based on the Big-Five Personality Test of Extroversion, Inquisitiveness, Accommodation, Emotional Stability and Orderliness. For each group based on personality type, the frequency of each self-reported academic emotion of confidence, excitement, frustration and interest was recorded and two classifiers, kNN and C4.5, were trained for each personality type. The accuracy rate of the classifiers built using only data instances from those assessed to be "low" in "orderliness", as well as only from those assessed to be "high" in "orderliness", performed significantly better compared to the classifiers that were trained for all personality types combined. The experiments also revealed that for almost all the 5 personality types, the percentage of instances where the learners reported themselves to be confident or frustrated differed significantly depending on whether they were assessed as "low" or "high" in the five personality types. © 2011 IEEE.