International Journal on Advanced Science, Engineering and Information Technology, Vol. 9 (2019) No. 6, pages: 1944-1949, DOI:10.18517/ijaseit.9.6.10229

Predicting Students’ Course Performance Based on Learners’ Characteristics via Fuzzy Modelling Approach

Chee Siong Teh, Shu Hsien Lee, Mohamad Hardyman Barawi


Frequent assessment allows instructors to ensure students have met the course learning objectives. Due to lack of instructor-student interaction, most of the assessment feedbacks and early interventions are not carried out in the large class size. This study is to proposes a new way of assessing student course performance using a fuzzy modeling approach. The typical steps in designing a fuzzy expert system include specifying the problem, determining linguistic variables, defining fuzzy sets as well as obtaining and constructing fuzzy rules is deployed. An educational expert is interviewed to define the relationship between the factors and student course performance. These steps help to determine the range of fuzzy sets and fuzzy rules in fuzzy reasoning. After the fuzzy assessing system has been built, it is used to compute the course performances of the students. The subject expert is asked to validate and verify system performance. Findings show that the developed system provides a faster and more effective way for instructors to assess the course performances of students in large class sizes.  However, in this study, the system is developed based on 150 historical student data and only a total of six factors related to course performance are considered. It is expected that considering more historical student data and adding more factors as the variables help to increase the accuracy of the system.


assessment feedback; fuzzy logic; fuzzy inference system; student modeling; intelligent tutoring system.

Viewed: 431 times (since abstract online)

cite this paper     download