Today, the rapid development of digitalization processes is unstable, primarily reflected in the specifics of using certain innovative practices. For example, the latter functions alongside traditional sources of data acquisition, processing, analysis, and generation (including textual data). In this context, neural network modeling should be positioned as a manifestation of the above processes, designed to dynamize modern linguistic research. To a greater extent, they (linguistic research) are focused on natural language processing, which, in turn, greatly impacts language analysis in general (from syntactic recognition to machine translation and text data generation) [1].