THE ANALYSIS OF LEARNING STYLE IN E-LEARNING ENVIRONMENT VIA WEB USAGE MINING

45475

The aim of this study is to perceive the learning styles of the students from the behaviours in e- learning environments with artificial intelligence models via web usage mining. Kolb s learning style classification and scale have been used for the learning styles of the students. 230 students participated in the study carried out by the education faculty students of Firat University. Data have been collected for learning styles before and after students started using the web-based system that has been prepared within the scope of the research. It has been found that students experience changes in learning styles while they are using e-learning environments. This suggests the need for systems that can take into account individual differences and the differences that students may experience with their interactions with the system. Artificial Neural Networks (ANN) and Decision Tree models have been applied in the studies to determine learning styles from students web usage behaviours. In the e-learning environment, the duration and frequency of visits to ten learning activities offered to students make up the entries of data mining models, and learning styles make up the output of the data mining models. When all the results obtained from Decision Tree and ANN models are evaluated, it is understood that ANN model is more successful than Decision Tree model. When the ANN model has been evaluated in terms of its algorithms, the Levenberg-Marquardt algorithm predicted learning styles of students with success of 91.7%.

987

E , learning environment , Learning style , Neural networks

Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi

1015

Abant İzzet Baysal Üniversitesi Eğitim Fakültesi Dergisi