Open AccessClustering student behavior based on quiz activities on moodle LMS to discover the relation with a final exam score
Author(s) -
Indra Waspada,
Nurdin Bahtiar,
Adi Wibowo
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1217/1/012118
Subject(s) - cluster analysis , centroid , cluster (spacecraft) , relation (database) , computer science , learning management , k means clustering , e learning , mathematics education , psychology , artificial intelligence , multimedia , world wide web , data mining , the internet , operating system
One strategy in using the Learning Management System (LMS) is to provide activities that can be accessed by students as learning material. The problem is that the teacher cannot see the students’ online activity directly. In this study, the students’ behavior was discovered from the e-learning log file on the quiz activity. Quis activities are performed six times and end with a final exam. Four attributes were proposed as the student behavior model, they are the frequency of repeating quiz work (freq), the duration on each quiz work (our), increased score when repeating the quiz (perf), and the best score of all attempts (best). The preferable clustering result using K-means algorithm was obtained with 2 (two) clusters. Each cluster was paired with the final exam score which shows that the percentage of failures in cluster_0 is 85%. This result corresponds to the centroid value in cluster_0 that is lower in all attributes compare to the cluster_1. It concludes to evidence that the activity of working on online quizzes has a relation with the final exam score. This information can be used for early intervention by the teacher to prevent the student from failing.