Data Mining for Exploring E-learning in a Computer Science Course Using Online Judging

Abstract

New teaching methods based on the students' learning process are being developed in the European Higher Education Area. Most of them are oriented to promote students' interest in the study and offer personalized feedback. On-line judging is a promising method for encouraging students’ participation in the elearning process. The great amount of data available in an on-line judging tool provides the possibility of exploring some of the most indicative attributes for learning programming concepts and techniques. In this paper, the results of programming activities carried out in a course on “Algorithms and Data Structures” has been used to identify the factors that affect the program correction, by using powerful data mining technologies taken from artificial intelligence domain. Concretely, our study uses a decision tree because it has been identified as the best predictor in some elearning domains. An overall accuracy of 60.1% in the prediction of the program correction was achieved with three input parameters (Programming language, Number of problem and Degree). In future work, we aim to analyze collaborative activities in order to identify the factors or predictor variables that affect workers’ performance in a global software development context