Frown Gives Game Away: Affect Sensitive Tutoring Systems for Elementary Mathematics*
Abdolhossein Sarrafzadeh, Chao Fan, Farbad Dadgostar, Samuel T.V. Alexander, Chris H. Messom · 2004
An important factor in the success of human one-to-one tutoring is the tutor's ability to identify and respond to affective cues. However, current Intelligent Tutoring Systems model students solely on their cognitive, and not affective, state. Automated facial expression analysis makes it possible for computers to identify affective cues by analysing images of learners' facial expressions. A new generation of Intelligent Tutoring Systems, Affective Tutoring Systems, is being developed at Massey University in New Zealand. An important component of this system is a facial expression analysis component that was implemented using a fiq approach. This &-q facial expression analysis approach and future directions are discussed in this paper. behaviour in real time and use this information to individualize interactions with the student. An Affective Tutoring System that is being developed in conjunction with the New Zealand Numeracy Project in the domain of counting and addition will be introduced. This paper introduces a facial expression analysis system that forms the basis of affective state detection. The implementation of an affective mathematics tutoring system is also discussed. A new fuzzy approach employed in the interpretation of facial expressions will also be presented. Based on previous studies (7), we discuss why we believe that without developing such tools first, affect sensitive tutoring systems are not possible.