AI-Tutor: Generating Tailored Remedial Questions and Answers Based on Cognitive Diagnostic Assessment

Developing AI-powered intelligent tutoring systems to facilitate adaptive learning has been a hot research topic due to that these systems can provide personalized learning guidance by taking individual learner's status and needs into consideration to improve their learning performance. Based on this perspective, this paper proposes a novel personalized adaptive tutoring system named AI-Tutor, which not only incorporates the basic functions of general tutoring systems that providing adaptive course learning to learners, but also can generate tailored remedial questions and answers based on cognitive diagnostic assessment. This unique and tailored tutoring service will potentially help learners master the deficient knowledge points in a much shorter time. This paper presents the preliminary research and development in implementing such an AI-powered tutor. First, it gives the system design with the function description of main modules and the operating procedure. Second, it identifies four main research problems behind this system and proposes approaches to solve these problems in developing the AI-Tutor.