DESIGN OF STUDENT SUCCESS PREDICTION APPLICATION IN ONLINE LEARNING USING FUZZY-KNN

Abstract

Effective evaluation of student performance is crucial. Hence, many kinds of techniques are used such as statistics, physical examination and currently data mining techniques to evaluate student performance. Data mining techniques as known as Educational Data Mining (EDM) collect, process, report and used to find the unseen patterns in the student dataset. EDM uses machine learning techniques to dig out useful data from multiple levels of meaningful hierarchy. Various data from intelligent computer tutors, classic computer based educational systems, online classes, academic data in educational institution, and standar assesment can be process for EDM. This led universities include open and distance learning (ODL) to collect large volume of student and learning data in their learning management systems (LMS). Students in ODL are relatively familiar with LMS and many learning activities such as number of accessing materials, student participation in discussion forum recorded in LMS. The processes of using EDM to improve the quality of educational policy maker with data-based models have become a challange that institutions of higher education face today. Therefore, this study aims to design applications that predict student performance in online learning using machine learning techniques based on EDM. The machine learning technique used in this research is Fuzzy-KNN. Testing using Fuzzy-KNN produces an accuracy of 92.5%.