HOW EDUCATIONAL DATA MINING CAN PREDICT STUDENTS’ ACADEMIC ACHIEVEMENT

Abstract

INTRODUCTION:Duta Wacana Christian University is located in Yogyakarta, Indonesia. This is one of the Christian Higher Education Institutions in Indonesia. Currently, Duta Wacana Christian University inspects problems with student's learning motivation, especially in Information System Departments. Information Systems students have average GPA during 2012 to 2015 is 3.1 (out of 4) and the study period 5.37 years (while the normal study period is 4 years according to the curriculum). It means that motivation among Information System Students is average. Not only GPA, study period is also a problem faced by Information System Department Head. Average study period is 5.37. This number is higher than the targeted study period which is 4 years according to the curriculum of Information System Department.Bahji, Lefdaoui, & Alami (2013) inferred that there is a decrease found in students' motivation to participate learning activity in the classroom. There are some factors that influence these findings which are as follows· The development of Information Technology· The usage of Information Technology· The usage of internet as a learning media toolFrom the table 1, it is observed that around 20% students are failed during their study. For sure, it is a big problem for Information System Department, School of Information Technology. Based on those problems, this research is conducted to study the correlation between academic achievement (as measured by GPA like Najafabadi, Najafabadi, and Farid-Rohani (2013) research did) to predict the percentage of successful students during their study.Table 1 also supports what Siang and Santoso (2016) study in its paper. Information System Department Students feel demotivated towards their study. Demotivated students can be seen with the high study period and many students are out during their study. Siang and Santoso (2016) found that students have willingness to increase their GPA up to 9.2 out of 10. This willingness is not supported with the eagerness to do the assignments and prepare for the exams. It means that students show less efforts to achieve their ambition to increase its GPA.Academic Achievement was a research topic for Najafabadi, Najafabadi, and Farid -Rohani (2013) whose study was about academic achievement, measured by GPA, for Shahid Bahesti University in Iran, found that there are some factors influence the academic achievement, such as Teaching - Evaluation, Learner, Environment, Family, Curriculum, and Teaching Knowledge.Clearly, academic performance is a critical factor to be taken into account. Underachievement is associated with a high dropout rate (Martinez, Karanik, Giovannini, and Pinto, 2015). Educational Data Mining is one of the solution that can predict and anticipate the higher number of student's underachievement and drop out. Martinez, et. al. (2015) found out an alternative solution to characterize the students into some certain groups. Thus, the profiling becomes a strategy of significant value when an action is taken to improve the performance of students.Following the introduction, literature review in Educational Data Mining is conducted. The next part is research methodology, followed by analysis and discussion. The final part from this study will be conclusion and further research suggestions.LITERATURE REVIEW:ACADEMIC ACHIEVEMENTS:Academic achievement is a measurement for students' academic performance. The measurement of students' academic performance is Grade Point Average (GPA). GPA can describe the students' performance during their study (Martinez et al., 2015). Bahji, et. al. (2013) showed that there are some factors that influence the learning process. Motivation and engagement are two factors that influence the process and its result. Learning can bring some changes in human knowledge and behavior, including social, psychology, and other kind of changes. …