Ontological Modeling of the Knowledge Base of Intellectual GIS of Digital Agriculture

Introduction. The field of digital agriculture requires effective management of agricultural resources based on intelligent analysis of heterogeneous spatiotemporal data collected from various sensor sources. Modern geographic information systems (GIS) allow collecting, processing and visualizing this data, but their capabilities for semantic information coordination and automated decision-making remain limited. The ontological approach provides systematization, structuring and interoperability of sensor data, formalization of domain knowledge, as well as intelligent extension of GIS functionality for solving applied tasks. The purpose of the paper. The research is aimed at developing an ontological model of the knowledge base of intelligent GIS of digital agriculture, which will provide a formalized representation, integration and processing of knowledge of a given subject area in the OWL format and will contribute to the automation of the analysis of agrotechnical processes, increasing the relevance of query results and optimizing decision-making based on taking into account the semantics of interoperable data. Results. The need for knowledge processing is substantiated to extract context, interpret and integrate heterogeneous data coming from different sources (agrodrones, autonomous tractors, cartographic services, etc.) and having different structures and levels of detail. The semantic and pragmatic aspects of the ontology are determined in the form of a mind map, which reflects the dimensions of the ontology in terms of formalization and detailing of information content and reuse of the ontology to solve new applied problems and extend the knowledge network. A formal ontological model of a knowledge base is proposed, which covers the key entities (categories) of digital agriculture (soils, crops, climatic factors, technical means and agro-technological operations) in two components – the four-component ontology containing interconnected sets: concepts (classes and subclasses), relations, interpretation functions, axioms, and a separate set of instances of defined concepts, which plays the role of a database with which the previous sets are linked. The proposed model was validated on test data in the Protege environment, which supports the representation of knowledge in OWL notation. A number of queries were generated for the constructed ontological knowledge base based on the SPARQL language. Conclusions. The developed ontological model of the knowledge base for the intelligent geoinformation system of digital agriculture provides semantic integration and interpretation of heterogeneous data, automation of decision-making and, as a result, increasing the efficiency of agricultural production, and also allows to create a flexible and adaptive system capable of evolution by extending the created model by integrating new concepts and relations between them. Further research on this topic involves the implementation of logical inference mechanisms within the model using SWRL rules to increase the level of automation of decision-making processes. Keywords: ontology, geographic information system, digital agriculture, Protege, OWL, RDF, SPARQL, knowledge base, semantics.