Mining VRSEC student learning behaviour in moodle system using datamining techniques

Predicting the performance of the students and helping them to improve their knowledge in subjects is one of the jobs of the educational universities. It is a laborious work to track many students in the universities. So, the universities started using content management systems to track the record of the student´s marks, grades and performance. Even then, the tutor have to evaluate manually to finalize the list of low grade students. As this is a problematic method, in this paper, the best method is explained for this purpose. There are many e-learning systems helping the institutions to evaluate their student´s skills. In this paper, the comparison for those existing ones are made and the better one is selected i.e., Moodle. Being an open source software, providing the users flexibility in updating the tasks and excelling in its security it is opted to store VRSEC student´s performance records. The further step is to analyze the students´ performance and predict the grades. For this, Rapid miner an open source tool is used. Primarily the records of the VRSEC students are stored in the Moodle and are then extracted into Rapid miner environment. Certain parameters are defined to complete the pre-processing concept. Then classification algorithm is applied to predict the grades of the students. By applying various classification algorithms, it is observed that decision tree algorithm gave greatest accuracy of 85% and with weighted mean of recall and precision as 75.00% and 89.63% respectively.