Penerapan Data Mining dalam Pengelompokan Uang Kuliah Tunggal (UKT) Menggunakan Metode K-Means Pada Universitas Jambi

Abstrak

UKT is a single tuition fee borned by each undergraduate student at a state university in Indonesia, to be paid every semester. At Jambi University, 8 UKT groups apply to students based on the level of the student's economic condition. Determining the UKT group for new students manually is less effective because it really depends on the assessment of the assessor, the criteria for economic conditions are quite a lot, the potential for subjectivity, and the amount of data is quite large. This study grouped new students into 8 UKT groups by applying data mining using the k-Means method. The k-Means method performs clustering based on the similarity of data on student economic conditions. K-Means analysis was performed using Rapidminer and SPSS tools. The results show that there are 7 variables/attributes of economic condition parameters that can be used in k-means analysis, including: total parent/guardian income including additional income, parent/guardian employment, electricity bills, parent status, land and building tax bills, housing conditions, and the number of dependents who are still at school. The results show that between the interpretation of the k-means analysis and the real data, there was a similarity in determining the UKT group above 50% in both Rapidminer and SPSS. Thus it can be concluded that the k-Means method can be applied to support decision making in determining UKT groups for students.