Automated Soil Testing Process using Combined Mining Patterns

Our proposed work presents an automated online system, which uses data mining techniques called combined mining patterns to predict the category of the analyzed soil datasets and provide suggestions of the crops which can be cultivated for better yield. The basic problem of predicting the crop yield is formalized by the classification rule, where Naive Bayes and K-Nearest Neighbor algorithms are used. The soil testing starts with the collection of a soil sample which is being collected from the agricultural field. The first and foremost principle of the soil testing process is that an agriculture field can be sampled in such a way that chemical analysis of the soil sample will accurately reflect the agricultural field's true nutrient status.