Resumen
The information space provides a large amount of destructive, stressful information, and restricts self-destructive behavior. The article considers the possibility of using automated analysis of destructive content of telegram channels to predict the risks of developing self-destructive behavior. A comprehensive approach has been implemented taking into account the theoretical analysis of socio-psychological factors of self-destructive behavior, unloading of the content of popular telegram channels. To collect information, the built-in messenger functions were used, a system for assessing the tonality and searching for messages on a given topic was developed for data processing. The study was conducted on the example of the analysis of 269 channels. The content of telegram channels was analyzed on 9 topics (the main types of self-destructive behavior, factors contributing to its development). As a result, the volume of destructive content for different categories of channels was calculated, significantly significant differences in the representation of destructive content on the main topics were revealed. A data collection model is also described, which can later be used to monitor and specify the risks of self-destructive behavior.