SoilBR-Onto: An Ontology for Soil Fertility Management and Classification in Brazil.

This article focuses on SoilBR-Onto, an ontology developed to enhance soil classification, fertilization, and liming recommendations for Brazilian agriculture based on laboratory soil analysis data. SoilBR-Onto incorporates 481 axioms and 145 semantic rules to enable precise inferences regarding soil texture, organic matter, phosphorus, potassium, and acidity correction, supporting decision-making in precision agriculture. The ontology is extensible from the AgroRDF framework and is implemented with tools such as the Pellet reasoner and the SoilViewer Mobile application, which provides users with accessible fertilization and liming guidance and visualizes soil deficiencies via heat maps. Validated with real soil samples, SoilBR-Onto adheres to FAIR (Findable, Accessible, Interoperable, Reusable) principles and is designed for adaptability to various crops beyond tobacco, including soybeans, wheat, and corn. Future work aims to expand its crop coverage, improve scalability, integrate machine learning for real-time inference, and enhance interoperability with other agricultural ontologies.