Knowledge-based expert system for agricultural land evaluation and crop selection
Knowledge based expert systems form a division of Artificial Intelligence systems that are becoming increasingly useful in decision support applications in domains that were once considered uniquely human in nature. Land evaluation and land use planning are two such domain areas. The role of decision support systems which are run by algorithms is found to be inadequate in land evaluation projects. Expert systems approach the problem by emulating human experts and having a continuously growing knowledge base. A land use expert system uses knowledge from various sources to evaluate a land unit’s resources to suggest an optimal land use or crop.
The current research paper proposes a knowledge-based approach to land evaluation for selection of suitable agricultural crops for a land unit using “Crop Advisor” – a Knowledge-based Spatial Decision Support System (KBSS) for land evaluation and crop selection. The “Crop Advisor” expert system utilises multiple knowledge bases to determine a suitable, optimal crop for a land unit considering the land unit’s inherent potentials and constraints. The expert system is powered primarily by human knowledge collected from land evaluation and crop experts. The knowledge bases constitute physical, economic, environmental and social factors that affect land use choices. The expert system makes use of Geographic Information System (GIS) tools to manage and handle spatial information that are required for the evaluation process. These powerful tools (ES and GIS) are used to assign crop choices that allow maximization of output complying with environmental values. It also considers economic feasibility of raising a crop by taking market price, cost of production, access to market, yield levels. The model suggests a strategic land use plan at regional level considering soil, climate and physiography which omits the non-feasible land use choices (crops). The “Crop Advisor” expert system then suggests in consultation with farmer (through a graphical user interface) a suitable agricultural crop that can be grown in a land unit with reasoning.