Classification of Students’ Achievement via Machine Learning by Using System Logs in Learning Management System

Abstract

During emergency remote teaching (ERT) process, factors affecting the achievement of students have changed. The purposes of this study are to determine the variables that affect the classification of students according to their course achievements in ERT during the pandemic process and to examine the classification performance of machine learning techniques. For these purposes, the logs from the learning management system were used. In the study, analyzes were carried out with various machine learning techniques and their performances were compared. As a result of the study, it was observed that Fisher’s Linear Discriminant Analysis was the best technique in classification according to F measure performance criteria. As another result, the most effective variable, in classifying students, is the average number of days logged into the system per month and week. It has been observed that total activity duration (min), total number of weeks and total number of page views during the semester are less influential factors. Accordingly, it could be suggested to check the monthly and weekly follow-up of the lectures instead of the total follow-ups per semester. In addition, students’ interaction patterns can be monitored with course tracking systems.

Proficiency of Teachers' Perceptions of Distance Education and Technology Usage Competencies: a Meta-Analysis Study

“…For this reason, world education ministries expect teachers to have the competence to use information and communication technologies effectively in the field of managing the learning and teaching process, in addition to the general competencies of the teaching profession (Agyei & Voogt, 2011;Topchyan & Zhang, 2014;Sangra et al,2012;Zhang et al, 2020). In addition, it is emphasized that teachers and teacher candidates responsible for the qualified education of individuals should have the characteristics that can renew and develop themselves in the face of rapidly changing and developing conditions as a requirement of the information society (Koyuncu et al, 2022). Teachers are not only responsible for implementing the curriculum in the education process (Altunoglu, 2017;Hodges & Cullen, 2020).…”

Uso de los algoritmos Machine Learning para analizar Moodle y los teléfonos inteligentes en el proceso educativo de la Física

“…Las instituciones educativas enfrentan el reto de cambiar las actividades de los cursos a través de las Tecnologías de la Información y Comunicación (TIC) y los modelos pedagógicos (GORMAZ-LOBOS; GALARCE-MIRANDA y KERSTEN, 2021;HEETHAL JAIPRAKASH, 2022;KOYUNCU;KILIC y GOKSUN, 2022). Los sistemas de gestión de aprendizaje, las herramientas en la nube Google, los sistemas de videoconferencia, los dispositivos móviles y las herramientas Web 2.0 permitieron la enseñanza y el aprendizaje durante el periodo del COVID-19 (KARAHISAR y UNLUER, 2022; KOYUNCU; KILIC y GOKSUN, 2022).…”

“…2 Revisión de Literatura 2.1 Ciencia de Datos Hoy en día, la Ciencia de Datos permite encontrar nuevos patrones para clasificar y evaluar la información (CHADAGA et al, 2021;KHANAL et al, 2020). De hecho, el uso de los algoritmos Aprendizaje Automático o Machine Learning ayuda a tomar una mejor decisión a través de la predicción de los fenómenos (CHADAGA et al, 2021;KHANAL et al, 2020;KOYUNCU;KILIC y GOKSUN, 2022).…”