Adaptive trainer for preparing students for math exams

Abstract

Due to the specificity of organizing the learning process in school classes, often students do not have enough time for a practice to solve tasks (like a math or in similar theme) under the supervision of a teacher. In this case, using the services of tutors or selftraining. In recent years, online educational services have become most popular for self-learning. This approach has some limitations, one of which is the inability to individually configure for a specific user. The presented project is aimed at automating the process of e-learning, regarding the acquisition of practical skills in solving mathematical tasks, determining the level of knowledge and reducing the duration of training by reducing the number of tasks depending on the level of training (adaptability). An adaptive testing approach has been implemented to satisfy these requirements. The created Markov models of adaptive testing became the basis for the development of an adaptive trainer for teaching nonformalized skills and abilities necessary for solving mathematical and other problems of sufficiently high complexity that require mastering both the standard technique of constructing reasoning and elements of creative thinking. A web service has been developed to demonstrate how it works. First of all, an adaptive trainer is considered the most effective in cases where it is necessary to order the knowledge and skills, such as to prepare for exams when solving non-formalized tasks (mathematical, technical, algorithmic, etc.). This service is not a substitute for a teacher, but rather complements the existing learning process, expanding the possibilities for self-study.