Employing Multi-Modal Affective Computing Techniques to Design the Personalized Human-Computer Interface of an Intelligent Tutoring System - An Example on Digital Arts Learning

Abstract

Some professional educators had indicated that emo tion is one of the important factors which may affect learning results. But, to combined Affective Computing and e- learning is still at the initial phase. Recently, t here are many projects establish a Intelligent Tutoring System to assist and enhance the learning effectiveness. Digital Arts, which are taken seriously by various teaching material design ers, are regarded as an important method to help promote education because of it's gr eat subjective and sensitive characters. So, base on this, this project will use Affective R ecognition to coordinate with Digital Arts, apply to the Intelligent Tutoring System to observe learner's emotion to judge if the learning process is smoothly and give feedback promptly and timely, hope to increase interactive between users and our system, to increase our syste m's usability, learning motivation and learning effectiveness. This project used two diffe rent methods to recognize users' emotions: Facial expression recognition and semanti c emotion recognition. Through this dual-mode mechanism, our system can enhance identif ied effectiveness and obtain more recognizable emotions. The best parts in this mecha nism is that we can aware students' learning status any time. Training of image process ing template is used for our Facial expression recognition and identify detected emotio nal keywords with syntactical logic is used for our semantic emotion recognition. Our syst em can analysis users' emotion and learning status to arrange proper teaching strategi es and courses, then our virtual teaching assistant could be a good channel of communication between users and our system, so we can expect better learning effectiveness from our learners.