Machine Learning based Smart Crop Recommendation System

Many challenges are faced by agricultural sector due to various dependent factors like changes in climatic conditions, variation in soil properties and farming practices. What to plant in their agricultural land is the common question for all farmers. Many farmers are failing to identify the crop that suits their agricultural land to get good yields. This results in following the same methods and crops what their ancestors were following. The crop yield always depends on multiple parameters like soil properties, climate conditions, water level etc. So if the farmers get suggestions regarding which crop to grow in their agricultural land would be of great help. The government is also helping farmers in many ways, one such scheme is soil testing. Soil testing is very important because through testing we can identify nutritional level, pH level and moisture level of soil which can optimize crop prediction. To address the above mentioned issues, this paper proposes a Crop Recommendation System(CRS) to help farmers based on machine learning algorithms to take wise decision to select a crop for their agricultural land. The CRS considers historical data, soil type, pH level, season, water level and amount of pesticides to provide crop recommendation for agricultural plots. The CRS system uses Naive Bayes algorithm to select the crop suitability. The CRS is a user friendly system that can be accessed through web or mobile. The user can give necessary input regarding their agricultural land and get crop recommendation which suits their farming land. The proposed system aims to improve crop yield, minimize resources and environmental impact.