Affective Virtual TA Project

Development of affective virtual TA service based on deep learning for foreign language education

This project aimed to develop an affective virtual teaching assistant (VTA) service platform using various deep learning methods for foreign language education. The dataset was collected from real flipped learning classes provided by Hankuk University of Foreign Studies. The developed service, named Edee, comprises four major functions: (1) Social Tutor, (2) Personal Tutor, (3) School Life Supporting TA, and (4) Supporting TA for Faculty. The function of Social Tutor is designed to foster social bonding among students through activities such as ice-breaking. The function of Personal Tutor aids foreign students in learning Korean, focusing on grammar, semantics, and conversational skills within situational scenarios in Korean real life. The function of School Life Supporting TA dynamically manages students' schedules. Lastly, the function of Supporting TA for Faculty provides professors with comprehensive student information (such as for understanding group mood among project teams) for each lecture to enhance the quality of instruction.