Indigenous Knowledge Crop Health Recommendation Expert System for the Upcoming Generations

In this paper, the researcher develops a mobile-based indigenous knowledge (IK) crop health expert system that receives information about the crop, weather, soil and environmental properties, and then provides a set of recommendations appropriate for the given input. The development of this system was motivated by three existing problems: (1) lack of systematic documentation and technological exploitation of the indigenous knowledge; (2) complications rising in crop development and assessment due to climate change and (3) degradation and loss of the IK from generation to generation. The system learns from the IK experts and simulates how these experts assess the health and development of their crops to maximize yield. This is achieved by using the machine learning clustering model that learns by grouping similar patterns of farmers’ knowledge data into clusters. Further, it will map the input object to the closest cluster and extracts all the recommendations inside the selected cluster. The system was evaluated with the help of IK experts and 72% accuracy was reached. The accuracy is deemed to increase for as long as more data is collected.