Share
Export Citation
Web-Based Student Academic Performance Evaluation System using Data Mining
Ilham A.
Aip Conference Proceedings
Abstract
This research aims to retrieve student academic performance profiles using data mining technology. The clustering process is implemented to the dataset of student academic performance of Informatics Undergraduate Program in Hasanuddin University, Indonesia, from the intake year 2008 to 2013. Using k-means algorithm with k=5, the algorithm retrieves two identical profiles. Profile 1 shows the student academic performance clusters based on GPA (Grade Point Average) and length of study while profile 2 shows the student academic performance clusters based on GPA, length of study, and total credits earned. The SSE (Sum Square Error) formula is used to measure the variation within a cluster in each profile. The smaller the SSE, the better the similarity within a cluster. The SSE of profile 1 (4.19628) is smaller than the SSE of profile 2 (6.34025). This result shows that profile 1 has better similarity within a cluster compared to profile 2.