Weekly Analysis of Moodle Log Data in RStudio for Future Use in Prediction

Abstract

This paper presents the results of the study to find in how many weeks the prediction/ classification can be done through a learning management system. The results are obtained by analyzing the effects of learners’ online presence and activities in Moodle on their grades at the early stages. Analyses are done by partitioning log data of a course in three years by time in terms of weeks. For this purpose we used RStudio and developed script to automatize the analysis. We found that starting from the third week of the lecture period online presence of the students becomes stable and classification can be done starting from that time and accordingly different material and assessment methods can be offered to the students in LMS.