In the current era of big data, information technology is developing quite rapidly, the most important data mining technology in information technology is also widely used, and now it is applied to the field of agricultural production, what can solve many problems such as agricultural production, fertilization and so on. In this paper, data mining technology is applied to the process of corn fertilization, because in corn production, effective and reasonable amount of fertilizer can make corn grow better, however, if there is no specific fertilization according to the soil properties of the corn, it will lead to the soil which needs fertility can not be with enough fertility, and the soil without fertility will be added more and more. In view of this problem, the soil planted with corn was graded and treated with different levels of soil, so as to achieve the purpose of effective utilization of soil fertility. In this paper, the DBSCAN algorithm in clustering analysis is used to classify the soil, the DBSCAN algorithm to this field have not been reported so far. By applying the nutrient balance method, the amount of soil fertilizer was calculated at each level, and the fertilizer was targeted according to the amount of fertilizer. Through the pilot application in Nong’an County of Jilin province Chen hometown, compared with the traditional fertilization results, Fertilizer reduced by 25%, corn production increased by about 15%, effectively reducing the input of chemical fertilizer and increasing the output of crops.