|
|
 |
| |
|
|
|
Utilizing the Educational Data Mining Techniques "Orange Technology" for Detecting Patterns and Predicting Academic Performance of University Students |
|
|
|
PP: 1415-1431 |
|
|
|
doi:10.18576/isl/120330
|
|
|
|
Author(s) |
|
|
|
A. S. Abdelmagid,
A. I. M. Qahmash,
|
|
|
|
Abstract |
|
|
| This study aimed at detecting the educational patterns and predicting the academic performance of university students through the “Orange” technology for data mining. To achieve this aim, a set of electronic courses taught to King Khalid University students through the Blackboard Learning Management System were selected. For knowledge detection, the "K-Means" clustering algorithm was used to detect patterns, while "Linear Regression, Random Forest, KNN, Tree, SVM" algorithms were used to predict students academic performance. The results indicated that the "K- Means" aggregation algorithm collected students scores in three main layers: the highest was in the first and second classes, while the lowest was in the third layer. As for predicting academic performance, the results indicated that students academic performance can be predicted through activities and quarterly tests for all courses except for one Course in which academic performance can be predicted through the semester tests only, and the quarterly activities do not contribute to predicting the students academic performance. The Linear Regression algorithm is the most contributing algorithm in predicting the academic performance of students, while the SVM algorithm is the least.
|
|
|
|
|
|
|
|
