Sistem Informasi Geografis Berbasis Web untuk Produksi Kedelai di Indonesia Menggunakan K-Means Clustering

Abstract

Information concerning areas that are potential for soybean production is not provided in an interactive information system that can be easily accessed by public. Nowadays, information about soybean production in all districts in Indonesia is displayed in the tabular form in the website http://aplikasi.deptan.go.id/bdsp. For that, in this study, a web-based geographic information system (GIS) was developed to manage and visualize soybean production in all districts in Indonesia. K-means clustering was performed to group data on production, productivity and harvested areas of soybean to make the coloring process become easier in GIS. The best clustering was obtained with 4 clusters and the sum square error (SSE) were 2.07, 1.53, and 0.0037 for data on production, productivity, and harvested areas of soybean, respectively. The GIS was developed using OpenGeo Suite 3.0 as the geospatial tool for managing and providing information about soybean production in Indonesia in forms of maps, tables, and graphs based on the results of k-means clustering.