<![CDATA[The Garut Regency Agriculture Office has a subsidized fertilizer distribution program, the amount of fertilizer distributed depends on the submission from the sub-district based on the area, then distributed in the disbursement period every 3 or 6 months. This program aims to help partner farmers or those recorded in the agriculture office to make plants included in the partner farmer program grow well. Based on the agricultural data owned, the subsidized fertilizer distribution program still cannot be improved, after monitoring or supervision activities on the data. Because the agriculture office only distributes fertilizers based on submissions from partner farmers, as well as the results of agricultural turnover in the form of nominal sales, making the data received cannot be evaluated whether the fertilizer distribution program is appropriate or not, and whether it meets the needs of farmers with a diversity of types of vegetables planted, which can provide good vegetable quality at harvest time and increase sales turnover. With the problems experienced by the Garut Regency Agriculture Office, a method is needed that can process agricultural data so that it can be one of the references in implementing the fertilizer distribution program from the agriculture office, to help partner farmers in fertilizing vegetable crops planted on land in the sub-districts planted, as well as policies that need to be evaluated from the characteristics of agricultural data that have been processed. The purpose of this research is to process agricultural data at the Garut Regency Agriculture Office into groups with their characteristics, by applying the K-Means algorithm. The research method is implemented with the Cross-Industry Standard Process for Data mining with the stages of Business understanding, Data understanding, Data Preparation, Modeling, Evaluation and Deployment. This study obtained clustering results with the characteristics of agricultural data based on clustering, the comparison data is on the land area of group 0 contains data with a land area below 20,183 m2, the amount of fertilizer below 404 sacks, and turnover below Rp 605,490,300, while Group 1 contains a land area above 64,875 m2, fertilizer above 1,298 sacks and turnover above Rp 1,946,262,600. The results of this study can be used as a reference for improvement in agricultural data with large land areas and the need for thousands of sacks of fertilizer use, such as divided into several groups, not only by one sub-district manager, but including the village government to manage the distribution of fertilizers because the land area, the amount of fertilizer and turnover is very high.]]>
