Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back The use of Local Generalizations in Developing Solutions in Spatial Situations S.L. BELYAKOV, A.V. ISAEV, D.O. GRIGORIEV Generating solutions in problematic spatial situations is considered one of the important tasks assigned to geospatial intelligence (GeoAI). Being an effective means of accumulating spatial data and decision-making experience, geoinformation systems do not provide the expected reliability and validity of decisions and recommendations. In conditions of uncertainty about the description and behavior of the real world, GeoAI systems lack generalized knowledge about local events and phenomena. The use of machine learning models gives good results on adequate training samples, which are not available in many applied fi elds. In this paper, we consider the solution to the problem of increasing reliability through the use of local generalizations that are created by experts in a retrospective analysis of actual situations. Local generalization refers to the representation of situations by a set of acceptable, unacceptable, and possible transformations. It is proposed to introduce a class of objects of local generalization into the information base of the geoinformation system. A conceptual model of the repeated application of local generalizations in comparing situations and developing solutions is described. Its distinctive feature is its connection to the topology of the area and the scheme of the problem situation. The results of an experiment on the expert assessment of the possibility of reuse of local generalizations are presented. Keywords: geoinformation modeling; spatial analysis; geospatial intelligence; use of experience. DOI: 10.25791/pribor.11.2025.1628 Pp. 19-25.