The architecture of educational platform based on machine learning

Abstract

The paper discusses the architecture of an educational platform based on machine learning (ML) algorithms. Recently, ML algorithms which is subpart of artificial intelligence (AI) have started using in online educational platform. This paper illustrates the architecture of education platform based on ML algorithms. ML algorithms use to identify students’ psychological state, to determine
the level of knowledge in subjects, to give educational materials suitable for the level of knowledge of the student, and to classify the exam result of a student into classes. The architecture in the paper is designed for the school, however, it is flexible. The architecture is able to apply for the other degrees of education. Thus, the platform based on ML is able to improve students autonomous learning
and each subject abilities. In 2024, 280 students from 6th to 11th grades were taught at private school in Tashkent City by the educational platform based on the architecture. Determining the students psychological condition, determining the students knowledge level, assessing the student’s knowledge from each subject, moving the student to the next level was performed with the help of SVM, ANN, Naive Bayes, and Decision tree machine learning techniques. The findings illustrate that ML algorithms are highly effective in teaching students.