Application Of The K-Means Clustering Method In Grouping Goat Productivity In A Web-Based Monitoring System

Abstract

Monitoring goat productivity at Pawit Mulya Farm Pacitan was previously hampered by manual data management and the lack of systematic productivity grouping. This study aims to develop a web-based monitoring system by implementing K-Means Clustering based on body weight, birth rate, and milk production. The study used a system development approach with the Waterfall method. The study population consisted of approximately 230 goats, while the sample consisted of productivity data that met the research variables. Data collection was carried out through observation, interviews, and literature studies. Analysis used Euclidean Distance, centroid updates, convergence testing, and Silhouette Score. The results of the study produced three productivity clusters: low, medium, and high, with a Silhouette Score of 0.3312, indicating sufficient clustering quality. White Box testing produced five valid independent paths with a 100% success rate. The system can be used as a means of monitoring and grouping productivity in a structured manner, although separation between clusters still needs to be optimized.