A Context Aware Recommender System for Predicting Crop Factors using LSTM

Agriculture has been one of the important industry sectors in the world as it is related to food security and survival of human race. Information and communication technologies have played a great role in increasing the food production and making agriculture a profitable business. There are various factors which directly or indirectly affect the quality and quantity of crops such as meteorological data, time, and raw seeds. With the advancement of internet of things and availability of sensors we can collect data related to most of the factors, which should be analyzed timely and effectively to make accurate decisions. In this paper, a Long Short-Term Memory (LSTM) based framework is proposed which takes multiple factors into consideration and predict crop factors related to a specific crop. The proposed framework uses feature extraction to extract important features for a crop as input and out-put both, then uses LSTM neural network to predict the specific crop fac-tors. The proposed framework is tested on real-time data collected from the internet and produced accuracy of more than 93%.