Segmentasi Siswa Berdasarkan Capaian Literasi dan Numerik Menggunakan Teknik Clustering

The Competency-Based National Assessment (CBNA) aims to measure students' basic competencies in literacy and numeracy as the foundation for learning at various levels of education. However, assessment results often show significant variations due to factors such as learning environment, teaching methods and socio-economic background, making it difficult for schools to design effective learning strategies. This study aims to map students based on literacy and numeracy achievement using clustering techniques with K-Means and Spectral Clustering algorithms. The data used is the 2023 National Assessment Public Report Card for SMA/SMK/MA/MAK levels. The analysis process includes pre-processing, exploratory analysis, application of clustering algorithm, and evaluation using Silhouette coefficient. The results showed that the K-Means algorithm with two clusters performed better (Silhouette coefficient 0.1901) than Spectral Clustering (-0.2218). The first cluster grouped students with low literacy and numeracy attainment, while the second cluster included students with higher attainment.