K-Means Clustering to Improve Interest Grouping Model For High School Students

Abstract

Informatics Engineering has a significant appeal to high school students in the digital era. However, differences in students' understanding of career prospects in this field affect their level of interest. This study aims to identify students' interest patterns using the K-Means Clustering algorithm as a basis for developing data-based strategies to increase the attractiveness of the major. This study used quantitative methods with primary data collected through questionnaires from 202 high school students. The variables analyzed include students' understanding of Informatics Engineering, interest in technology subjects, and aspirations to continue their studies in the field. The data was processed using RapidMiner software, through the stages of pre-processing, data transformation, and model evaluation. Davies-Bouldin Index (DBI) was used to determine the best number of clusters, with cluster trials (k) from 2 to 10. The results showed the best DBI value at k=2 with a score of 0.527. Two clusters were formed: Cluster 0 (uninterested students) with 96 students and Cluster 1 (interested students) with 106 students. Interested students generally have a better understanding of career prospects in technology, while less interested students need additional education to increase their interest. This research shows the importance of a data-driven approach in understanding student needs. For students with low interest, an upstream program is needed.