Intelligent Analysis of Educational Data in the Adaptive E-Learning Environment

Abstract

In the process of adaptive learning, individual selection of tasks for students training is of great importance. The selection of training tasks in the adaptive e-learning environment can be performed automatically using methods of intelligent analysis of educational data. The article presents the research results of key intelligent problem of individual selection of training tasks, which is the prediction of the task difficulty for a student, taking into account student performance. The results of the prediction will allow us to perform an individual selection of such training tasks in which the student will receive the best learning results. The article suggests and justifies the authors' approach to solving the problem of predicting the difficulty of training tasks based on the automatic classification of pairs "student-task" using artificial intelligence methods. The opportunities of trainable models of neural networks and decision trees are investigated, the problem of forming a training sample based on the data of the e-learning system is discussed. The results of an experiment on actual educational data are presented.