Crops Recommendation System Model Using Weather Attributes, Soil Properties, and Crops Prices

Abstract

Agriculture is a respectful and predominant occupation in India. Approximately, 60% population works in this industry contributing to 18% GDP of the country. Weather and soil conditions are key factors to be considered for selecting a crop to be farmed. Even after growing a crop successfully, it should provide profitable returns to the farmer. Traditionally, farmers choose a crop to be harvested from past experiences and some wise men advices. But there are chances that predictions may go wrong assuming incorrect past experiences since human brains are good at analyzing rather than remembering. There were many articles and research works which tried to recommend crops based on either weather or soil properties. But a methodology which includes both of them along with consideration of crop prices will be of more useful to farmers. So, a proper system using weather attributes, taking soil inputs, and analyzing profit earned could help farmers in taking a right decision of crops to be farmed. Using this model, we propose a system that can recommend highly profitable crops to farmers. For predicting weather attributes like temperature and rainfall, we can use past data of a region applied to ARIMA model. Logistic regression can be used for classification of data based on weather and soil conditions. And for predicting profit that can be attained, we can use investments and returns data of past years applied to ARIMA model.