A Crop Selection Framework using K Nearest Neighbour for IoT Semantic Interoperability Applications

In agrosystems, the soil plays a determining role through the multitude of ecosystem services that it provides. Agriculture exploits natural spaces and resources for useful production to humans. Soil is the primary source of nutrients for the plants. The roots absorb nutrient ions from soil and water. Additional nutrients must be mixed in the soil before the cultivation begins. Use of excess chemicals has adverse effects both on the crop and the environment. Optimal amount of nutrient requirement must be met for maximum yield. This paper presents a crop selection criteria based on K-nearest neighbour algorithm. The proposed method used farm inputs like soil type, climate, micronutrients, macronutrients, water source etc. to select the best crop suited for the farm. The algorithm also provides the amount of deficient nutrients required for the crop production. This methodology can be integrated into a semantic interoperability framework for Internet of Things (IoT) based applications. The proposed method accurately estimated the crop needs and the corresponding quantity of the nutrient necessary and sufficient to achieve a production objective defined by the yield.