Integrated introspective case-based reasoning for intelligent tutoring systems

Many intelligent tutoring systems (ITSs) have been developed, deployed, assessed, and proven to facilitate learning. However, most of these systems do not generally adapt to new circumstances, do not self-evaluate and self-configure their own strategies, and do not monitor the usage history of the learning content being delivered or presented to the students. These shortcomings force ITS developers to often spend much development time in manual revision and finetuning of the learning and instructional contents of an ITS. In this paper, we describe an intelligent agent that delivers learning material adaptively to different students, factoring in the usage history of the learning materials and student profiles as observed by the agent. Student-tutor interaction includes the activities of going through learning material, such as a topical tutorial, a set of examples, and a set of problems. Our assumption is that our agent will be able to capture and utilize these student activities as the primer to select the appropriate examples or problems to administer to the student. Using an integrated introspective case-based reasoning approach, our agent further learns from its experience and refines its reasoning process - including the instructional strategies - to adapt to student needs. Moreover, our agent monitors the usage history of the learning materials to improve its performance. We have built an end-to-end ITS using an agent powered by this integrated introspective case-based reasoning engine. We have deployed the ITS in a CS course. Results indicate that the ITS was able to learn to deliver more appropriate examples and problems to the students.