Learning Performance Improvement Through Participation in Online Seminar: Machine Learning Analysis

Abstract

Learning performance is related to students’ learning activities during a learning process. Their learning behavior could lead to successful course accomplishment or not, to better or worse final marks. Seminar practices have their impact on development of some students’ competences like: topics analysis, discussion and presentation and the planned tasks concern learning performance. In online environment, the seminars could be organized in the form of different learning scenarios and it depends on the functional and technical features of the organized educational environment as well as on the course goal. In this paper an investigation and analysis of students’ participation in online seminars is conducted with aim to understand the dependence between their learning performance, online tasks realization and final results. eLearning informatics gives possibilities for usage contemporary methods for research and learning analytics as one of them is machine learning. Machine learning algorithms are utilized to group students according to their learning behavior and final outcome. The created analytical models could be in support of educators and students to improve their educational activities. The accuracy of machine learning algorithms is evaluated to find the best model according to collected data during one semester.