SenAgriPreci: Improving Precision Agricultural Yields Through a Crop Recommender System

Agriculture represents an important pillar in the global economy. It plays a preponderant role in social life. Africa alone represents 30% of the earth’s arable land, or 30 million km2 of usable agricultural land. This sector employs more than 65% of the active Senegalese population. However, inappropriate weather or climate change have a deep effect on agricultural production or crop yields. Consequently, the conformance of precision agriculture requires effective decisions by stakeholders in the agricultural field. This can be achieved through an innovative recommender system. A recommender system filters data using different algorithms and suggests the most relevant items to users in a particular domain using relevant entities. To address the above issue, we propose a recommender system for farmers in Senegal called SENAgriPreci which enables users to make good and precise agricultural decisions using various parameters (Example: pH, crop type, soil type, temperature etc.). Using relevant datasets, namely: SAED and CRNA, we conducted benchmarking experiments to validate our proposed SENAgriPreci recommendation methodology. Experimental results in terms of Root Mean Square Error (RSME) validates that of our proposed recommendation method is more favorable and outperforms other related and contemporary recommendation methods in relation to accuracy and precise crop recommendations.