Behavior Pattern Mining during the Evaluation Phase in an e-Learning Course
Authors
Abstract
There is a broad range of products available for e-learning which can be used in course curriculum at University level. While e-learning in education is well established, there are a few attempts to extract information in its evaluation phase. We look at a specific part of an e-learning designed course favoring students' evaluation phase for extraction of behavior pattern mining. Our approach uses neural networks (NN) and Support Vector Machines (SVM) to build prediction models able to track student's behavior. The data sets were obtained from student's logs in a Moodle designed Course of Discrete Structures of the Informatics Engineering Bachelor at the University of Coimbra. The results show the model is able to successfully predict students' final outcome while bringing useful feedback during course learning.
Keywords
e-learning, data mining techniques, course evaluation
Subject
Data mining; course evaluation
Conference
ICEE 2007 - International Conference on Engineering Education, September 2007