There is often a need to analyze unstructured data when assessing the risk of emergency situations. Traditional analysis methods may not take into account the ambiguity of information, which makes them insufficiently effective for risk assessment. The article proposes the use of a modified hierarchy process analysis method using fuzzy logic, which allows for more effective consideration of uncertainties and subjective assessments in the process of analyzing emergency risks. In addition, such methods allow for consideration of not only quantitative indicators, but also qualitative ones. This, in turn, can lead to more informed decisions in the field of risk management and increased preparedness for various situations. The integration of technologies for working with unstructured data in the process of assessing emergency risks not only increases the accuracy of forecasting, but also allows for adapting management strategies to changing conditions.

Keywords: artificial intelligent systems, unstructured data, risk assessment, classical hierarchy analysis method, modified hierarchy analysis method, fuzzy logical inference system

The article is devoted to applied issues of improving regional security management processes through the development of methods for analyzing data on emergency situations. In order to identify patterns in the occurrence of emergency situations, multidimensional methods of processing statistical data were used. A multidimensional classification of data in the field of emergency situations based on fuzzy logic is proposed. The classification was performed using a fuzzy inference system with clear membership functions. As statistical data, data on emergency situations of a man-made, natural and biological-social nature that occurred in the federal districts of Russia in 2020, including data on dead and saved people, were considered. An analysis of data samples on regional emergency situations was carried out according to 5 criteria, and clustering of regions was carried out.

Keywords: emergency situations, fuzzy multidimensional clustering, fuzzy logic, fuzzy inference system, computer program, mathematical model, forecasting, decision making