Making the most from virtual learning environments captivates researchers, enhancing the learning experience and reducing the withdrawal rate. In that regard, this article presents a framework for a withdrawal prediction model for the data of the Open University, one of the largest distance-learning institutions. The main contributions of this work cover two main aspects: relational-to-tabular data transformation and data mining for withdrawal prediction. This main steps of the process are: (1) tackling the unbalanced data issue using the SMOTE algorithm; (2) voting over seven different features' selection algorithms; and (3) learning different classifiers for withdrawal prediction. The experimental study demonstrates that the decision trees exhibit better performance in terms of the F-measure value compared to the other tested models. Furthermore, the data balancing and feature selection processes show a crucial role for guiding the predictive model towards a reliable module.
Nowadays, the virtual learning environment has become an ideal tool for professional self-development and bringing courses for various learner audiences across the world. There is currently an increasing interest in researching the topic of learner dropout and low completion in distance learning, with one of the main concerns being elevated rates of occurrence. Therefore, the early prediction of learner withdrawal has become a major challenge, as well as identifying the factors, which contribute to this increasingly occurring phenomenon. In that regard, this manuscript presents a framework for withdrawal prediction model for the data from The Open University, one of the largest distance learning institutions. For that purpose, we start by pre-processing the dataset and tackling the challenge of discretization process and unbalanced data. Secondly, this paper identifies the semantical issues of raw data by introducing new behavioural indicators. Finally, we reckon on machine learning algorithms for withdrawal prediction model to understand the lack of learners' commitment at an early stage.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.