Analysis of the Russian narrative trend models in the information space

Currently, the global information space is filled with narratives. They are varying in purpose and level of stratification. The narratives of the occupying forces have been aimed at different target audiences since the beginning of the Russian-Ukrainian war. These narratives pose a threat to the information security of the state, society, and the individual. Timely detection of the main messages that are subordinate to Russian narratives can serve as indicators of the information operation. It is quite important to understand the concept of the sequence of reactions to independent events, that may be mistakenly perceived as containing messages. The analysis should contain a combination of theoretical and situational explanations of the revealed regularities and the specified outbursts. It is recommended to determine the limitations that will be applied in the study for the formation of data input. Namely, these are three basic narratives, the statistics must be significant enough for a considerable length of time and they are designed for three different target audiences. According to accepted restrictions it is popular to use special software to collect statistics, which allows analysts to extract data from the content of the information space following specified requests. This is usually possible due to the use of methods of artificial intelligence and machine learning. Access to the received sets of semantic and metric data (data sets) allows the use of mathematical methods of analysis. One of the common approaches is to find the regression function by comparing the obtained approximation functions. Such a comparison of the obtained mathematical models can be performed considering the nature of the phenomenon, accuracy indicators and the adequacy of the behavior of the regression model during its extrapolation. Accordingly, it is possible to determine the main dynamic-behavioral features of the phases of the Russian information operation due to a rough comparison of the convergence of the regression curves and the standard one.