CGSC 2001 Lecture Notes - Lecture 10: Intelligent Tutoring System, Directed Acyclic Graph, Bayesian Network
Document Summary
Back in the 80"s, bloom was exploring how different contexts influence what students learn. Bloom compared: classroom instruction: one teacher for a group of students, individualized human tutoring: a tutor works with one or two students in private sessions. Findings: human tutoring resulted in significantly higher learning gains: human tutoring resulted in performance increase of 2 grade levels over classroom instruction. Unfortunately, human tutoring is not a scalable approach cannot provide private tutor for each student. Challenge: to find a means of instruction that approximates the benefits of human tutoring and can reach many students: field of intelligent tutoring systems. Computer applications that aim to support student learning by providing individualized help and instruction, tailored to a given student"s needs. Its are based on the assumption that human tutors are effective because they adapt to student"s needs: offer tailored feedback, including elaborated help that targets student misconceptions.