Abstract:
We live in an e-learning era, where the fast growth of e-learning around the world is inspiring many educational and business institutions. The rapid growth of e-learning has changed the traditional learning behavior and presented a new situation to both lecturers and students. Lecturers find it harder to guide students to select suitable learning materials due to overdosed learning resources. In the other hand learners spend more time on browsing and filtering learning resources to identify the suitable materials, rather than focusing on learning all the materials. Hence, the learning material recommendation is an essential requirement for e-learning system, which would take the e-learning, to the next level.
Though e-learning brings many benefits in comparison with the conventional learning paradigm, with the rapid increase of learning contents on the web, the e- learning lacks proper feedback or the guidance, which is a key in the traditional teaching process to identify the relevant resources. The students who do not have knowledge to find out the most suitable resources, links and references for their studies and the assignments, may waste most of their time on browsing in search of the relevant material, without any guidance. Some of the “good students” may indirectly act as good guides to their fellow friends. The fellow average learners could follow the methods adapted by good students in learning or accessing relevant learning material. They may refer to the mostly accessed “learning materials” by the “good learners”. The lack of feedback in e-learning systems could be overcome by evolving an intelligent feedback system that would recommend the heavily accessed resources by the good students to the average students and that is the aim of this project.
The research collected the Moodle log data of the courses that contain a rich set of electronic course contents from Department of Computer Science and Engineering, University of Moratuwa. This audited data of the collected courses have been used to construct “student classification” and “resource recommendation” models. The former classifies the students as “Good students” and “Average Students” and the latter recommends the “mostly accessed resources” by good students.
The results show that the resources that are mostly accessed by good students are more probable to be recommendable resources. The learning resource recommendation framework helps the average students who fail to choose the relevant materials for their studies from the heaps of learning resources. By making the website interactive, the communication between faculty and students could be made more effective, hence that would promote active learning.
